1 @chapter Filtering Introduction
2 @c man begin FILTERING INTRODUCTION
4 Filtering in FFmpeg is enabled through the libavfilter library.
6 In libavfilter, a filter can have multiple inputs and multiple
8 To illustrate the sorts of things that are possible, we consider the
13 input --> split ---------------------> overlay --> output
16 +-----> crop --> vflip -------+
19 This filtergraph splits the input stream in two streams, then sends one
20 stream through the crop filter and the vflip filter, before merging it
21 back with the other stream by overlaying it on top. You can use the
22 following command to achieve this:
25 ffmpeg -i INPUT -vf "split [main][tmp]; [tmp] crop=iw:ih/2:0:0, vflip [flip]; [main][flip] overlay=0:H/2" OUTPUT
28 The result will be that the top half of the video is mirrored
29 onto the bottom half of the output video.
31 Filters in the same linear chain are separated by commas, and distinct
32 linear chains of filters are separated by semicolons. In our example,
33 @var{crop,vflip} are in one linear chain, @var{split} and
34 @var{overlay} are separately in another. The points where the linear
35 chains join are labelled by names enclosed in square brackets. In the
36 example, the split filter generates two outputs that are associated to
37 the labels @var{[main]} and @var{[tmp]}.
39 The stream sent to the second output of @var{split}, labelled as
40 @var{[tmp]}, is processed through the @var{crop} filter, which crops
41 away the lower half part of the video, and then vertically flipped. The
42 @var{overlay} filter takes in input the first unchanged output of the
43 split filter (which was labelled as @var{[main]}), and overlay on its
44 lower half the output generated by the @var{crop,vflip} filterchain.
46 Some filters take in input a list of parameters: they are specified
47 after the filter name and an equal sign, and are separated from each other
50 There exist so-called @var{source filters} that do not have an
51 audio/video input, and @var{sink filters} that will not have audio/video
54 @c man end FILTERING INTRODUCTION
57 @c man begin GRAPH2DOT
59 The @file{graph2dot} program included in the FFmpeg @file{tools}
60 directory can be used to parse a filtergraph description and issue a
61 corresponding textual representation in the dot language.
68 to see how to use @file{graph2dot}.
70 You can then pass the dot description to the @file{dot} program (from
71 the graphviz suite of programs) and obtain a graphical representation
74 For example the sequence of commands:
76 echo @var{GRAPH_DESCRIPTION} | \
77 tools/graph2dot -o graph.tmp && \
78 dot -Tpng graph.tmp -o graph.png && \
82 can be used to create and display an image representing the graph
83 described by the @var{GRAPH_DESCRIPTION} string. Note that this string must be
84 a complete self-contained graph, with its inputs and outputs explicitly defined.
85 For example if your command line is of the form:
87 ffmpeg -i infile -vf scale=640:360 outfile
89 your @var{GRAPH_DESCRIPTION} string will need to be of the form:
91 nullsrc,scale=640:360,nullsink
93 you may also need to set the @var{nullsrc} parameters and add a @var{format}
94 filter in order to simulate a specific input file.
98 @chapter Filtergraph description
99 @c man begin FILTERGRAPH DESCRIPTION
101 A filtergraph is a directed graph of connected filters. It can contain
102 cycles, and there can be multiple links between a pair of
103 filters. Each link has one input pad on one side connecting it to one
104 filter from which it takes its input, and one output pad on the other
105 side connecting it to one filter accepting its output.
107 Each filter in a filtergraph is an instance of a filter class
108 registered in the application, which defines the features and the
109 number of input and output pads of the filter.
111 A filter with no input pads is called a "source", and a filter with no
112 output pads is called a "sink".
114 @anchor{Filtergraph syntax}
115 @section Filtergraph syntax
117 A filtergraph has a textual representation, which is recognized by the
118 @option{-filter}/@option{-vf}/@option{-af} and
119 @option{-filter_complex} options in @command{ffmpeg} and
120 @option{-vf}/@option{-af} in @command{ffplay}, and by the
121 @code{avfilter_graph_parse_ptr()} function defined in
122 @file{libavfilter/avfilter.h}.
124 A filterchain consists of a sequence of connected filters, each one
125 connected to the previous one in the sequence. A filterchain is
126 represented by a list of ","-separated filter descriptions.
128 A filtergraph consists of a sequence of filterchains. A sequence of
129 filterchains is represented by a list of ";"-separated filterchain
132 A filter is represented by a string of the form:
133 [@var{in_link_1}]...[@var{in_link_N}]@var{filter_name}=@var{arguments}[@var{out_link_1}]...[@var{out_link_M}]
135 @var{filter_name} is the name of the filter class of which the
136 described filter is an instance of, and has to be the name of one of
137 the filter classes registered in the program.
138 The name of the filter class is optionally followed by a string
141 @var{arguments} is a string which contains the parameters used to
142 initialize the filter instance. It may have one of two forms:
146 A ':'-separated list of @var{key=value} pairs.
149 A ':'-separated list of @var{value}. In this case, the keys are assumed to be
150 the option names in the order they are declared. E.g. the @code{fade} filter
151 declares three options in this order -- @option{type}, @option{start_frame} and
152 @option{nb_frames}. Then the parameter list @var{in:0:30} means that the value
153 @var{in} is assigned to the option @option{type}, @var{0} to
154 @option{start_frame} and @var{30} to @option{nb_frames}.
157 A ':'-separated list of mixed direct @var{value} and long @var{key=value}
158 pairs. The direct @var{value} must precede the @var{key=value} pairs, and
159 follow the same constraints order of the previous point. The following
160 @var{key=value} pairs can be set in any preferred order.
164 If the option value itself is a list of items (e.g. the @code{format} filter
165 takes a list of pixel formats), the items in the list are usually separated by
168 The list of arguments can be quoted using the character @samp{'} as initial
169 and ending mark, and the character @samp{\} for escaping the characters
170 within the quoted text; otherwise the argument string is considered
171 terminated when the next special character (belonging to the set
172 @samp{[]=;,}) is encountered.
174 The name and arguments of the filter are optionally preceded and
175 followed by a list of link labels.
176 A link label allows one to name a link and associate it to a filter output
177 or input pad. The preceding labels @var{in_link_1}
178 ... @var{in_link_N}, are associated to the filter input pads,
179 the following labels @var{out_link_1} ... @var{out_link_M}, are
180 associated to the output pads.
182 When two link labels with the same name are found in the
183 filtergraph, a link between the corresponding input and output pad is
186 If an output pad is not labelled, it is linked by default to the first
187 unlabelled input pad of the next filter in the filterchain.
188 For example in the filterchain
190 nullsrc, split[L1], [L2]overlay, nullsink
192 the split filter instance has two output pads, and the overlay filter
193 instance two input pads. The first output pad of split is labelled
194 "L1", the first input pad of overlay is labelled "L2", and the second
195 output pad of split is linked to the second input pad of overlay,
196 which are both unlabelled.
198 In a filter description, if the input label of the first filter is not
199 specified, "in" is assumed; if the output label of the last filter is not
200 specified, "out" is assumed.
202 In a complete filterchain all the unlabelled filter input and output
203 pads must be connected. A filtergraph is considered valid if all the
204 filter input and output pads of all the filterchains are connected.
206 Libavfilter will automatically insert @ref{scale} filters where format
207 conversion is required. It is possible to specify swscale flags
208 for those automatically inserted scalers by prepending
209 @code{sws_flags=@var{flags};}
210 to the filtergraph description.
212 Here is a BNF description of the filtergraph syntax:
214 @var{NAME} ::= sequence of alphanumeric characters and '_'
215 @var{LINKLABEL} ::= "[" @var{NAME} "]"
216 @var{LINKLABELS} ::= @var{LINKLABEL} [@var{LINKLABELS}]
217 @var{FILTER_ARGUMENTS} ::= sequence of chars (possibly quoted)
218 @var{FILTER} ::= [@var{LINKLABELS}] @var{NAME} ["=" @var{FILTER_ARGUMENTS}] [@var{LINKLABELS}]
219 @var{FILTERCHAIN} ::= @var{FILTER} [,@var{FILTERCHAIN}]
220 @var{FILTERGRAPH} ::= [sws_flags=@var{flags};] @var{FILTERCHAIN} [;@var{FILTERGRAPH}]
223 @section Notes on filtergraph escaping
225 Filtergraph description composition entails several levels of
226 escaping. See @ref{quoting_and_escaping,,the "Quoting and escaping"
227 section in the ffmpeg-utils(1) manual,ffmpeg-utils} for more
228 information about the employed escaping procedure.
230 A first level escaping affects the content of each filter option
231 value, which may contain the special character @code{:} used to
232 separate values, or one of the escaping characters @code{\'}.
234 A second level escaping affects the whole filter description, which
235 may contain the escaping characters @code{\'} or the special
236 characters @code{[],;} used by the filtergraph description.
238 Finally, when you specify a filtergraph on a shell commandline, you
239 need to perform a third level escaping for the shell special
240 characters contained within it.
242 For example, consider the following string to be embedded in
243 the @ref{drawtext} filter description @option{text} value:
245 this is a 'string': may contain one, or more, special characters
248 This string contains the @code{'} special escaping character, and the
249 @code{:} special character, so it needs to be escaped in this way:
251 text=this is a \'string\'\: may contain one, or more, special characters
254 A second level of escaping is required when embedding the filter
255 description in a filtergraph description, in order to escape all the
256 filtergraph special characters. Thus the example above becomes:
258 drawtext=text=this is a \\\'string\\\'\\: may contain one\, or more\, special characters
260 (note that in addition to the @code{\'} escaping special characters,
261 also @code{,} needs to be escaped).
263 Finally an additional level of escaping is needed when writing the
264 filtergraph description in a shell command, which depends on the
265 escaping rules of the adopted shell. For example, assuming that
266 @code{\} is special and needs to be escaped with another @code{\}, the
267 previous string will finally result in:
269 -vf "drawtext=text=this is a \\\\\\'string\\\\\\'\\\\: may contain one\\, or more\\, special characters"
272 @chapter Timeline editing
274 Some filters support a generic @option{enable} option. For the filters
275 supporting timeline editing, this option can be set to an expression which is
276 evaluated before sending a frame to the filter. If the evaluation is non-zero,
277 the filter will be enabled, otherwise the frame will be sent unchanged to the
278 next filter in the filtergraph.
280 The expression accepts the following values:
283 timestamp expressed in seconds, NAN if the input timestamp is unknown
286 sequential number of the input frame, starting from 0
289 the position in the file of the input frame, NAN if unknown
293 width and height of the input frame if video
296 Additionally, these filters support an @option{enable} command that can be used
297 to re-define the expression.
299 Like any other filtering option, the @option{enable} option follows the same
302 For example, to enable a blur filter (@ref{smartblur}) from 10 seconds to 3
303 minutes, and a @ref{curves} filter starting at 3 seconds:
305 smartblur = enable='between(t,10,3*60)',
306 curves = enable='gte(t,3)' : preset=cross_process
309 See @code{ffmpeg -filters} to view which filters have timeline support.
311 @c man end FILTERGRAPH DESCRIPTION
313 @chapter Audio Filters
314 @c man begin AUDIO FILTERS
316 When you configure your FFmpeg build, you can disable any of the
317 existing filters using @code{--disable-filters}.
318 The configure output will show the audio filters included in your
321 Below is a description of the currently available audio filters.
325 A compressor is mainly used to reduce the dynamic range of a signal.
326 Especially modern music is mostly compressed at a high ratio to
327 improve the overall loudness. It's done to get the highest attention
328 of a listener, "fatten" the sound and bring more "power" to the track.
329 If a signal is compressed too much it may sound dull or "dead"
330 afterwards or it may start to "pump" (which could be a powerful effect
331 but can also destroy a track completely).
332 The right compression is the key to reach a professional sound and is
333 the high art of mixing and mastering. Because of its complex settings
334 it may take a long time to get the right feeling for this kind of effect.
336 Compression is done by detecting the volume above a chosen level
337 @code{threshold} and dividing it by the factor set with @code{ratio}.
338 So if you set the threshold to -12dB and your signal reaches -6dB a ratio
339 of 2:1 will result in a signal at -9dB. Because an exact manipulation of
340 the signal would cause distortion of the waveform the reduction can be
341 levelled over the time. This is done by setting "Attack" and "Release".
342 @code{attack} determines how long the signal has to rise above the threshold
343 before any reduction will occur and @code{release} sets the time the signal
344 has to fall below the threshold to reduce the reduction again. Shorter signals
345 than the chosen attack time will be left untouched.
346 The overall reduction of the signal can be made up afterwards with the
347 @code{makeup} setting. So compressing the peaks of a signal about 6dB and
348 raising the makeup to this level results in a signal twice as loud than the
349 source. To gain a softer entry in the compression the @code{knee} flattens the
350 hard edge at the threshold in the range of the chosen decibels.
352 The filter accepts the following options:
356 Set input gain. Default is 1. Range is between 0.015625 and 64.
359 If a signal of second stream rises above this level it will affect the gain
360 reduction of the first stream.
361 By default it is 0.125. Range is between 0.00097563 and 1.
364 Set a ratio by which the signal is reduced. 1:2 means that if the level
365 rose 4dB above the threshold, it will be only 2dB above after the reduction.
366 Default is 2. Range is between 1 and 20.
369 Amount of milliseconds the signal has to rise above the threshold before gain
370 reduction starts. Default is 20. Range is between 0.01 and 2000.
373 Amount of milliseconds the signal has to fall below the threshold before
374 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
377 Set the amount by how much signal will be amplified after processing.
378 Default is 2. Range is from 1 and 64.
381 Curve the sharp knee around the threshold to enter gain reduction more softly.
382 Default is 2.82843. Range is between 1 and 8.
385 Choose if the @code{average} level between all channels of input stream
386 or the louder(@code{maximum}) channel of input stream affects the
387 reduction. Default is @code{average}.
390 Should the exact signal be taken in case of @code{peak} or an RMS one in case
391 of @code{rms}. Default is @code{rms} which is mostly smoother.
394 How much to use compressed signal in output. Default is 1.
395 Range is between 0 and 1.
400 Copy the input audio source unchanged to the output. This is mainly useful for
405 Apply cross fade from one input audio stream to another input audio stream.
406 The cross fade is applied for specified duration near the end of first stream.
408 The filter accepts the following options:
412 Specify the number of samples for which the cross fade effect has to last.
413 At the end of the cross fade effect the first input audio will be completely
414 silent. Default is 44100.
417 Specify the duration of the cross fade effect. See
418 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
419 for the accepted syntax.
420 By default the duration is determined by @var{nb_samples}.
421 If set this option is used instead of @var{nb_samples}.
424 Should first stream end overlap with second stream start. Default is enabled.
427 Set curve for cross fade transition for first stream.
430 Set curve for cross fade transition for second stream.
432 For description of available curve types see @ref{afade} filter description.
439 Cross fade from one input to another:
441 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:c1=exp:c2=exp output.flac
445 Cross fade from one input to another but without overlapping:
447 ffmpeg -i first.flac -i second.flac -filter_complex acrossfade=d=10:o=0:c1=exp:c2=exp output.flac
453 Reduce audio bit resolution.
455 This filter is bit crusher with enhanced functionality. A bit crusher
456 is used to audibly reduce number of bits an audio signal is sampled
457 with. This doesn't change the bit depth at all, it just produces the
458 effect. Material reduced in bit depth sounds more harsh and "digital".
459 This filter is able to even round to continuous values instead of discrete
461 Additionally it has a D/C offset which results in different crushing of
462 the lower and the upper half of the signal.
463 An Anti-Aliasing setting is able to produce "softer" crushing sounds.
465 Another feature of this filter is the logarithmic mode.
466 This setting switches from linear distances between bits to logarithmic ones.
467 The result is a much more "natural" sounding crusher which doesn't gate low
468 signals for example. The human ear has a logarithmic perception, too
469 so this kind of crushing is much more pleasant.
470 Logarithmic crushing is also able to get anti-aliased.
472 The filter accepts the following options:
488 Can be linear: @code{lin} or logarithmic: @code{log}.
497 Set sample reduction.
500 Enable LFO. By default disabled.
511 Delay one or more audio channels.
513 Samples in delayed channel are filled with silence.
515 The filter accepts the following option:
519 Set list of delays in milliseconds for each channel separated by '|'.
520 At least one delay greater than 0 should be provided.
521 Unused delays will be silently ignored. If number of given delays is
522 smaller than number of channels all remaining channels will not be delayed.
523 If you want to delay exact number of samples, append 'S' to number.
530 Delay first channel by 1.5 seconds, the third channel by 0.5 seconds and leave
531 the second channel (and any other channels that may be present) unchanged.
537 Delay second channel by 500 samples, the third channel by 700 samples and leave
538 the first channel (and any other channels that may be present) unchanged.
546 Apply echoing to the input audio.
548 Echoes are reflected sound and can occur naturally amongst mountains
549 (and sometimes large buildings) when talking or shouting; digital echo
550 effects emulate this behaviour and are often used to help fill out the
551 sound of a single instrument or vocal. The time difference between the
552 original signal and the reflection is the @code{delay}, and the
553 loudness of the reflected signal is the @code{decay}.
554 Multiple echoes can have different delays and decays.
556 A description of the accepted parameters follows.
560 Set input gain of reflected signal. Default is @code{0.6}.
563 Set output gain of reflected signal. Default is @code{0.3}.
566 Set list of time intervals in milliseconds between original signal and reflections
567 separated by '|'. Allowed range for each @code{delay} is @code{(0 - 90000.0]}.
568 Default is @code{1000}.
571 Set list of loudnesses of reflected signals separated by '|'.
572 Allowed range for each @code{decay} is @code{(0 - 1.0]}.
573 Default is @code{0.5}.
580 Make it sound as if there are twice as many instruments as are actually playing:
582 aecho=0.8:0.88:60:0.4
586 If delay is very short, then it sound like a (metallic) robot playing music:
592 A longer delay will sound like an open air concert in the mountains:
594 aecho=0.8:0.9:1000:0.3
598 Same as above but with one more mountain:
600 aecho=0.8:0.9:1000|1800:0.3|0.25
605 Audio emphasis filter creates or restores material directly taken from LPs or
606 emphased CDs with different filter curves. E.g. to store music on vinyl the
607 signal has to be altered by a filter first to even out the disadvantages of
608 this recording medium.
609 Once the material is played back the inverse filter has to be applied to
610 restore the distortion of the frequency response.
612 The filter accepts the following options:
622 Set filter mode. For restoring material use @code{reproduction} mode, otherwise
623 use @code{production} mode. Default is @code{reproduction} mode.
626 Set filter type. Selects medium. Can be one of the following:
638 select Compact Disc (CD).
644 select 50µs (FM-KF).
646 select 75µs (FM-KF).
652 Modify an audio signal according to the specified expressions.
654 This filter accepts one or more expressions (one for each channel),
655 which are evaluated and used to modify a corresponding audio signal.
657 It accepts the following parameters:
661 Set the '|'-separated expressions list for each separate channel. If
662 the number of input channels is greater than the number of
663 expressions, the last specified expression is used for the remaining
666 @item channel_layout, c
667 Set output channel layout. If not specified, the channel layout is
668 specified by the number of expressions. If set to @samp{same}, it will
669 use by default the same input channel layout.
672 Each expression in @var{exprs} can contain the following constants and functions:
676 channel number of the current expression
679 number of the evaluated sample, starting from 0
685 time of the evaluated sample expressed in seconds
688 @item nb_out_channels
689 input and output number of channels
692 the value of input channel with number @var{CH}
695 Note: this filter is slow. For faster processing you should use a
704 aeval=val(ch)/2:c=same
708 Invert phase of the second channel:
717 Apply fade-in/out effect to input audio.
719 A description of the accepted parameters follows.
723 Specify the effect type, can be either @code{in} for fade-in, or
724 @code{out} for a fade-out effect. Default is @code{in}.
726 @item start_sample, ss
727 Specify the number of the start sample for starting to apply the fade
728 effect. Default is 0.
731 Specify the number of samples for which the fade effect has to last. At
732 the end of the fade-in effect the output audio will have the same
733 volume as the input audio, at the end of the fade-out transition
734 the output audio will be silence. Default is 44100.
737 Specify the start time of the fade effect. Default is 0.
738 The value must be specified as a time duration; see
739 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
740 for the accepted syntax.
741 If set this option is used instead of @var{start_sample}.
744 Specify the duration of the fade effect. See
745 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
746 for the accepted syntax.
747 At the end of the fade-in effect the output audio will have the same
748 volume as the input audio, at the end of the fade-out transition
749 the output audio will be silence.
750 By default the duration is determined by @var{nb_samples}.
751 If set this option is used instead of @var{nb_samples}.
754 Set curve for fade transition.
756 It accepts the following values:
759 select triangular, linear slope (default)
761 select quarter of sine wave
763 select half of sine wave
765 select exponential sine wave
769 select inverted parabola
783 select inverted quarter of sine wave
785 select inverted half of sine wave
787 select double-exponential seat
789 select double-exponential sigmoid
797 Fade in first 15 seconds of audio:
803 Fade out last 25 seconds of a 900 seconds audio:
805 afade=t=out:st=875:d=25
810 Apply arbitrary expressions to samples in frequency domain.
814 Set frequency domain real expression for each separate channel separated
815 by '|'. Default is "1".
816 If the number of input channels is greater than the number of
817 expressions, the last specified expression is used for the remaining
821 Set frequency domain imaginary expression for each separate channel
822 separated by '|'. If not set, @var{real} option is used.
824 Each expression in @var{real} and @var{imag} can contain the following
832 current frequency bin number
835 number of available bins
838 channel number of the current expression
850 It accepts the following values:
866 Default is @code{w4096}
869 Set window function. Default is @code{hann}.
872 Set window overlap. If set to 1, the recommended overlap for selected
873 window function will be picked. Default is @code{0.75}.
880 Leave almost only low frequencies in audio:
882 afftfilt="1-clip((b/nb)*b,0,1)"
888 Apply an arbitrary Frequency Impulse Response filter.
890 This filter is designed for applying long FIR filters,
891 up to 30 seconds long.
893 It can be used as component for digital crossover filters,
894 room equalization, cross talk cancellation, wavefield synthesis,
895 auralization, ambiophonics and ambisonics.
897 This filter uses second stream as FIR coefficients.
898 If second stream holds single channel, it will be used
899 for all input channels in first stream, otherwise
900 number of channels in second stream must be same as
901 number of channels in first stream.
903 It accepts the following parameters:
907 Set dry gain. This sets input gain.
910 Set wet gain. This sets final output gain.
913 Set Impulse Response filter length. Default is 1, which means whole IR is processed.
916 Enable applying gain measured from power of IR.
923 Apply reverb to stream using mono IR file as second input, complete command using ffmpeg:
925 ffmpeg -i input.wav -i middle_tunnel_1way_mono.wav -lavfi afir output.wav
932 Set output format constraints for the input audio. The framework will
933 negotiate the most appropriate format to minimize conversions.
935 It accepts the following parameters:
939 A '|'-separated list of requested sample formats.
942 A '|'-separated list of requested sample rates.
944 @item channel_layouts
945 A '|'-separated list of requested channel layouts.
947 See @ref{channel layout syntax,,the Channel Layout section in the ffmpeg-utils(1) manual,ffmpeg-utils}
948 for the required syntax.
951 If a parameter is omitted, all values are allowed.
953 Force the output to either unsigned 8-bit or signed 16-bit stereo
955 aformat=sample_fmts=u8|s16:channel_layouts=stereo
960 A gate is mainly used to reduce lower parts of a signal. This kind of signal
961 processing reduces disturbing noise between useful signals.
963 Gating is done by detecting the volume below a chosen level @var{threshold}
964 and dividing it by the factor set with @var{ratio}. The bottom of the noise
965 floor is set via @var{range}. Because an exact manipulation of the signal
966 would cause distortion of the waveform the reduction can be levelled over
967 time. This is done by setting @var{attack} and @var{release}.
969 @var{attack} determines how long the signal has to fall below the threshold
970 before any reduction will occur and @var{release} sets the time the signal
971 has to rise above the threshold to reduce the reduction again.
972 Shorter signals than the chosen attack time will be left untouched.
976 Set input level before filtering.
977 Default is 1. Allowed range is from 0.015625 to 64.
980 Set the level of gain reduction when the signal is below the threshold.
981 Default is 0.06125. Allowed range is from 0 to 1.
984 If a signal rises above this level the gain reduction is released.
985 Default is 0.125. Allowed range is from 0 to 1.
988 Set a ratio by which the signal is reduced.
989 Default is 2. Allowed range is from 1 to 9000.
992 Amount of milliseconds the signal has to rise above the threshold before gain
994 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
997 Amount of milliseconds the signal has to fall below the threshold before the
998 reduction is increased again. Default is 250 milliseconds.
999 Allowed range is from 0.01 to 9000.
1002 Set amount of amplification of signal after processing.
1003 Default is 1. Allowed range is from 1 to 64.
1006 Curve the sharp knee around the threshold to enter gain reduction more softly.
1007 Default is 2.828427125. Allowed range is from 1 to 8.
1010 Choose if exact signal should be taken for detection or an RMS like one.
1011 Default is @code{rms}. Can be @code{peak} or @code{rms}.
1014 Choose if the average level between all channels or the louder channel affects
1016 Default is @code{average}. Can be @code{average} or @code{maximum}.
1021 The limiter prevents an input signal from rising over a desired threshold.
1022 This limiter uses lookahead technology to prevent your signal from distorting.
1023 It means that there is a small delay after the signal is processed. Keep in mind
1024 that the delay it produces is the attack time you set.
1026 The filter accepts the following options:
1030 Set input gain. Default is 1.
1033 Set output gain. Default is 1.
1036 Don't let signals above this level pass the limiter. Default is 1.
1039 The limiter will reach its attenuation level in this amount of time in
1040 milliseconds. Default is 5 milliseconds.
1043 Come back from limiting to attenuation 1.0 in this amount of milliseconds.
1044 Default is 50 milliseconds.
1047 When gain reduction is always needed ASC takes care of releasing to an
1048 average reduction level rather than reaching a reduction of 0 in the release
1052 Select how much the release time is affected by ASC, 0 means nearly no changes
1053 in release time while 1 produces higher release times.
1056 Auto level output signal. Default is enabled.
1057 This normalizes audio back to 0dB if enabled.
1060 Depending on picked setting it is recommended to upsample input 2x or 4x times
1061 with @ref{aresample} before applying this filter.
1065 Apply a two-pole all-pass filter with central frequency (in Hz)
1066 @var{frequency}, and filter-width @var{width}.
1067 An all-pass filter changes the audio's frequency to phase relationship
1068 without changing its frequency to amplitude relationship.
1070 The filter accepts the following options:
1074 Set frequency in Hz.
1077 Set method to specify band-width of filter.
1090 Specify the band-width of a filter in width_type units.
1093 Specify which channels to filter, by default all available are filtered.
1100 The filter accepts the following options:
1104 Set the number of loops.
1107 Set maximal number of samples.
1110 Set first sample of loop.
1116 Merge two or more audio streams into a single multi-channel stream.
1118 The filter accepts the following options:
1123 Set the number of inputs. Default is 2.
1127 If the channel layouts of the inputs are disjoint, and therefore compatible,
1128 the channel layout of the output will be set accordingly and the channels
1129 will be reordered as necessary. If the channel layouts of the inputs are not
1130 disjoint, the output will have all the channels of the first input then all
1131 the channels of the second input, in that order, and the channel layout of
1132 the output will be the default value corresponding to the total number of
1135 For example, if the first input is in 2.1 (FL+FR+LF) and the second input
1136 is FC+BL+BR, then the output will be in 5.1, with the channels in the
1137 following order: a1, a2, b1, a3, b2, b3 (a1 is the first channel of the
1138 first input, b1 is the first channel of the second input).
1140 On the other hand, if both input are in stereo, the output channels will be
1141 in the default order: a1, a2, b1, b2, and the channel layout will be
1142 arbitrarily set to 4.0, which may or may not be the expected value.
1144 All inputs must have the same sample rate, and format.
1146 If inputs do not have the same duration, the output will stop with the
1149 @subsection Examples
1153 Merge two mono files into a stereo stream:
1155 amovie=left.wav [l] ; amovie=right.mp3 [r] ; [l] [r] amerge
1159 Multiple merges assuming 1 video stream and 6 audio streams in @file{input.mkv}:
1161 ffmpeg -i input.mkv -filter_complex "[0:1][0:2][0:3][0:4][0:5][0:6] amerge=inputs=6" -c:a pcm_s16le output.mkv
1167 Mixes multiple audio inputs into a single output.
1169 Note that this filter only supports float samples (the @var{amerge}
1170 and @var{pan} audio filters support many formats). If the @var{amix}
1171 input has integer samples then @ref{aresample} will be automatically
1172 inserted to perform the conversion to float samples.
1176 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex amix=inputs=3:duration=first:dropout_transition=3 OUTPUT
1178 will mix 3 input audio streams to a single output with the same duration as the
1179 first input and a dropout transition time of 3 seconds.
1181 It accepts the following parameters:
1185 The number of inputs. If unspecified, it defaults to 2.
1188 How to determine the end-of-stream.
1192 The duration of the longest input. (default)
1195 The duration of the shortest input.
1198 The duration of the first input.
1202 @item dropout_transition
1203 The transition time, in seconds, for volume renormalization when an input
1204 stream ends. The default value is 2 seconds.
1208 @section anequalizer
1210 High-order parametric multiband equalizer for each channel.
1212 It accepts the following parameters:
1216 This option string is in format:
1217 "c@var{chn} f=@var{cf} w=@var{w} g=@var{g} t=@var{f} | ..."
1218 Each equalizer band is separated by '|'.
1222 Set channel number to which equalization will be applied.
1223 If input doesn't have that channel the entry is ignored.
1226 Set central frequency for band.
1227 If input doesn't have that frequency the entry is ignored.
1230 Set band width in hertz.
1233 Set band gain in dB.
1236 Set filter type for band, optional, can be:
1240 Butterworth, this is default.
1251 With this option activated frequency response of anequalizer is displayed
1255 Set video stream size. Only useful if curves option is activated.
1258 Set max gain that will be displayed. Only useful if curves option is activated.
1259 Setting this to a reasonable value makes it possible to display gain which is derived from
1260 neighbour bands which are too close to each other and thus produce higher gain
1261 when both are activated.
1264 Set frequency scale used to draw frequency response in video output.
1265 Can be linear or logarithmic. Default is logarithmic.
1268 Set color for each channel curve which is going to be displayed in video stream.
1269 This is list of color names separated by space or by '|'.
1270 Unrecognised or missing colors will be replaced by white color.
1273 @subsection Examples
1277 Lower gain by 10 of central frequency 200Hz and width 100 Hz
1278 for first 2 channels using Chebyshev type 1 filter:
1280 anequalizer=c0 f=200 w=100 g=-10 t=1|c1 f=200 w=100 g=-10 t=1
1284 @subsection Commands
1286 This filter supports the following commands:
1289 Alter existing filter parameters.
1290 Syntax for the commands is : "@var{fN}|f=@var{freq}|w=@var{width}|g=@var{gain}"
1292 @var{fN} is existing filter number, starting from 0, if no such filter is available
1294 @var{freq} set new frequency parameter.
1295 @var{width} set new width parameter in herz.
1296 @var{gain} set new gain parameter in dB.
1298 Full filter invocation with asendcmd may look like this:
1299 asendcmd=c='4.0 anequalizer change 0|f=200|w=50|g=1',anequalizer=...
1304 Pass the audio source unchanged to the output.
1308 Pad the end of an audio stream with silence.
1310 This can be used together with @command{ffmpeg} @option{-shortest} to
1311 extend audio streams to the same length as the video stream.
1313 A description of the accepted options follows.
1317 Set silence packet size. Default value is 4096.
1320 Set the number of samples of silence to add to the end. After the
1321 value is reached, the stream is terminated. This option is mutually
1322 exclusive with @option{whole_len}.
1325 Set the minimum total number of samples in the output audio stream. If
1326 the value is longer than the input audio length, silence is added to
1327 the end, until the value is reached. This option is mutually exclusive
1328 with @option{pad_len}.
1331 If neither the @option{pad_len} nor the @option{whole_len} option is
1332 set, the filter will add silence to the end of the input stream
1335 @subsection Examples
1339 Add 1024 samples of silence to the end of the input:
1345 Make sure the audio output will contain at least 10000 samples, pad
1346 the input with silence if required:
1348 apad=whole_len=10000
1352 Use @command{ffmpeg} to pad the audio input with silence, so that the
1353 video stream will always result the shortest and will be converted
1354 until the end in the output file when using the @option{shortest}
1357 ffmpeg -i VIDEO -i AUDIO -filter_complex "[1:0]apad" -shortest OUTPUT
1362 Add a phasing effect to the input audio.
1364 A phaser filter creates series of peaks and troughs in the frequency spectrum.
1365 The position of the peaks and troughs are modulated so that they vary over time, creating a sweeping effect.
1367 A description of the accepted parameters follows.
1371 Set input gain. Default is 0.4.
1374 Set output gain. Default is 0.74
1377 Set delay in milliseconds. Default is 3.0.
1380 Set decay. Default is 0.4.
1383 Set modulation speed in Hz. Default is 0.5.
1386 Set modulation type. Default is triangular.
1388 It accepts the following values:
1397 Audio pulsator is something between an autopanner and a tremolo.
1398 But it can produce funny stereo effects as well. Pulsator changes the volume
1399 of the left and right channel based on a LFO (low frequency oscillator) with
1400 different waveforms and shifted phases.
1401 This filter have the ability to define an offset between left and right
1402 channel. An offset of 0 means that both LFO shapes match each other.
1403 The left and right channel are altered equally - a conventional tremolo.
1404 An offset of 50% means that the shape of the right channel is exactly shifted
1405 in phase (or moved backwards about half of the frequency) - pulsator acts as
1406 an autopanner. At 1 both curves match again. Every setting in between moves the
1407 phase shift gapless between all stages and produces some "bypassing" sounds with
1408 sine and triangle waveforms. The more you set the offset near 1 (starting from
1409 the 0.5) the faster the signal passes from the left to the right speaker.
1411 The filter accepts the following options:
1415 Set input gain. By default it is 1. Range is [0.015625 - 64].
1418 Set output gain. By default it is 1. Range is [0.015625 - 64].
1421 Set waveform shape the LFO will use. Can be one of: sine, triangle, square,
1422 sawup or sawdown. Default is sine.
1425 Set modulation. Define how much of original signal is affected by the LFO.
1428 Set left channel offset. Default is 0. Allowed range is [0 - 1].
1431 Set right channel offset. Default is 0.5. Allowed range is [0 - 1].
1434 Set pulse width. Default is 1. Allowed range is [0 - 2].
1437 Set possible timing mode. Can be one of: bpm, ms or hz. Default is hz.
1440 Set bpm. Default is 120. Allowed range is [30 - 300]. Only used if timing
1444 Set ms. Default is 500. Allowed range is [10 - 2000]. Only used if timing
1448 Set frequency in Hz. Default is 2. Allowed range is [0.01 - 100]. Only used
1449 if timing is set to hz.
1455 Resample the input audio to the specified parameters, using the
1456 libswresample library. If none are specified then the filter will
1457 automatically convert between its input and output.
1459 This filter is also able to stretch/squeeze the audio data to make it match
1460 the timestamps or to inject silence / cut out audio to make it match the
1461 timestamps, do a combination of both or do neither.
1463 The filter accepts the syntax
1464 [@var{sample_rate}:]@var{resampler_options}, where @var{sample_rate}
1465 expresses a sample rate and @var{resampler_options} is a list of
1466 @var{key}=@var{value} pairs, separated by ":". See the
1467 @ref{Resampler Options,,the "Resampler Options" section in the
1468 ffmpeg-resampler(1) manual,ffmpeg-resampler}
1469 for the complete list of supported options.
1471 @subsection Examples
1475 Resample the input audio to 44100Hz:
1481 Stretch/squeeze samples to the given timestamps, with a maximum of 1000
1482 samples per second compensation:
1484 aresample=async=1000
1490 Reverse an audio clip.
1492 Warning: This filter requires memory to buffer the entire clip, so trimming
1495 @subsection Examples
1499 Take the first 5 seconds of a clip, and reverse it.
1501 atrim=end=5,areverse
1505 @section asetnsamples
1507 Set the number of samples per each output audio frame.
1509 The last output packet may contain a different number of samples, as
1510 the filter will flush all the remaining samples when the input audio
1513 The filter accepts the following options:
1517 @item nb_out_samples, n
1518 Set the number of frames per each output audio frame. The number is
1519 intended as the number of samples @emph{per each channel}.
1520 Default value is 1024.
1523 If set to 1, the filter will pad the last audio frame with zeroes, so
1524 that the last frame will contain the same number of samples as the
1525 previous ones. Default value is 1.
1528 For example, to set the number of per-frame samples to 1234 and
1529 disable padding for the last frame, use:
1531 asetnsamples=n=1234:p=0
1536 Set the sample rate without altering the PCM data.
1537 This will result in a change of speed and pitch.
1539 The filter accepts the following options:
1542 @item sample_rate, r
1543 Set the output sample rate. Default is 44100 Hz.
1548 Show a line containing various information for each input audio frame.
1549 The input audio is not modified.
1551 The shown line contains a sequence of key/value pairs of the form
1552 @var{key}:@var{value}.
1554 The following values are shown in the output:
1558 The (sequential) number of the input frame, starting from 0.
1561 The presentation timestamp of the input frame, in time base units; the time base
1562 depends on the filter input pad, and is usually 1/@var{sample_rate}.
1565 The presentation timestamp of the input frame in seconds.
1568 position of the frame in the input stream, -1 if this information in
1569 unavailable and/or meaningless (for example in case of synthetic audio)
1578 The sample rate for the audio frame.
1581 The number of samples (per channel) in the frame.
1584 The Adler-32 checksum (printed in hexadecimal) of the audio data. For planar
1585 audio, the data is treated as if all the planes were concatenated.
1587 @item plane_checksums
1588 A list of Adler-32 checksums for each data plane.
1594 Display time domain statistical information about the audio channels.
1595 Statistics are calculated and displayed for each audio channel and,
1596 where applicable, an overall figure is also given.
1598 It accepts the following option:
1601 Short window length in seconds, used for peak and trough RMS measurement.
1602 Default is @code{0.05} (50 milliseconds). Allowed range is @code{[0.1 - 10]}.
1606 Set metadata injection. All the metadata keys are prefixed with @code{lavfi.astats.X},
1607 where @code{X} is channel number starting from 1 or string @code{Overall}. Default is
1610 Available keys for each channel are:
1643 For example full key look like this @code{lavfi.astats.1.DC_offset} or
1644 this @code{lavfi.astats.Overall.Peak_count}.
1646 For description what each key means read below.
1649 Set number of frame after which stats are going to be recalculated.
1650 Default is disabled.
1653 A description of each shown parameter follows:
1657 Mean amplitude displacement from zero.
1660 Minimal sample level.
1663 Maximal sample level.
1665 @item Min difference
1666 Minimal difference between two consecutive samples.
1668 @item Max difference
1669 Maximal difference between two consecutive samples.
1671 @item Mean difference
1672 Mean difference between two consecutive samples.
1673 The average of each difference between two consecutive samples.
1675 @item RMS difference
1676 Root Mean Square difference between two consecutive samples.
1680 Standard peak and RMS level measured in dBFS.
1684 Peak and trough values for RMS level measured over a short window.
1687 Standard ratio of peak to RMS level (note: not in dB).
1690 Flatness (i.e. consecutive samples with the same value) of the signal at its peak levels
1691 (i.e. either @var{Min level} or @var{Max level}).
1694 Number of occasions (not the number of samples) that the signal attained either
1695 @var{Min level} or @var{Max level}.
1698 Overall bit depth of audio. Number of bits used for each sample.
1705 The filter accepts exactly one parameter, the audio tempo. If not
1706 specified then the filter will assume nominal 1.0 tempo. Tempo must
1707 be in the [0.5, 2.0] range.
1709 @subsection Examples
1713 Slow down audio to 80% tempo:
1719 To speed up audio to 125% tempo:
1727 Trim the input so that the output contains one continuous subpart of the input.
1729 It accepts the following parameters:
1732 Timestamp (in seconds) of the start of the section to keep. I.e. the audio
1733 sample with the timestamp @var{start} will be the first sample in the output.
1736 Specify time of the first audio sample that will be dropped, i.e. the
1737 audio sample immediately preceding the one with the timestamp @var{end} will be
1738 the last sample in the output.
1741 Same as @var{start}, except this option sets the start timestamp in samples
1745 Same as @var{end}, except this option sets the end timestamp in samples instead
1749 The maximum duration of the output in seconds.
1752 The number of the first sample that should be output.
1755 The number of the first sample that should be dropped.
1758 @option{start}, @option{end}, and @option{duration} are expressed as time
1759 duration specifications; see
1760 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}.
1762 Note that the first two sets of the start/end options and the @option{duration}
1763 option look at the frame timestamp, while the _sample options simply count the
1764 samples that pass through the filter. So start/end_pts and start/end_sample will
1765 give different results when the timestamps are wrong, inexact or do not start at
1766 zero. Also note that this filter does not modify the timestamps. If you wish
1767 to have the output timestamps start at zero, insert the asetpts filter after the
1770 If multiple start or end options are set, this filter tries to be greedy and
1771 keep all samples that match at least one of the specified constraints. To keep
1772 only the part that matches all the constraints at once, chain multiple atrim
1775 The defaults are such that all the input is kept. So it is possible to set e.g.
1776 just the end values to keep everything before the specified time.
1781 Drop everything except the second minute of input:
1783 ffmpeg -i INPUT -af atrim=60:120
1787 Keep only the first 1000 samples:
1789 ffmpeg -i INPUT -af atrim=end_sample=1000
1796 Apply a two-pole Butterworth band-pass filter with central
1797 frequency @var{frequency}, and (3dB-point) band-width width.
1798 The @var{csg} option selects a constant skirt gain (peak gain = Q)
1799 instead of the default: constant 0dB peak gain.
1800 The filter roll off at 6dB per octave (20dB per decade).
1802 The filter accepts the following options:
1806 Set the filter's central frequency. Default is @code{3000}.
1809 Constant skirt gain if set to 1. Defaults to 0.
1812 Set method to specify band-width of filter.
1825 Specify the band-width of a filter in width_type units.
1828 Specify which channels to filter, by default all available are filtered.
1833 Apply a two-pole Butterworth band-reject filter with central
1834 frequency @var{frequency}, and (3dB-point) band-width @var{width}.
1835 The filter roll off at 6dB per octave (20dB per decade).
1837 The filter accepts the following options:
1841 Set the filter's central frequency. Default is @code{3000}.
1844 Set method to specify band-width of filter.
1857 Specify the band-width of a filter in width_type units.
1860 Specify which channels to filter, by default all available are filtered.
1865 Boost or cut the bass (lower) frequencies of the audio using a two-pole
1866 shelving filter with a response similar to that of a standard
1867 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
1869 The filter accepts the following options:
1873 Give the gain at 0 Hz. Its useful range is about -20
1874 (for a large cut) to +20 (for a large boost).
1875 Beware of clipping when using a positive gain.
1878 Set the filter's central frequency and so can be used
1879 to extend or reduce the frequency range to be boosted or cut.
1880 The default value is @code{100} Hz.
1883 Set method to specify band-width of filter.
1896 Determine how steep is the filter's shelf transition.
1899 Specify which channels to filter, by default all available are filtered.
1904 Apply a biquad IIR filter with the given coefficients.
1905 Where @var{b0}, @var{b1}, @var{b2} and @var{a0}, @var{a1}, @var{a2}
1906 are the numerator and denominator coefficients respectively.
1907 and @var{channels}, @var{c} specify which channels to filter, by default all
1908 available are filtered.
1911 Bauer stereo to binaural transformation, which improves headphone listening of
1912 stereo audio records.
1914 It accepts the following parameters:
1918 Pre-defined crossfeed level.
1922 Default level (fcut=700, feed=50).
1925 Chu Moy circuit (fcut=700, feed=60).
1928 Jan Meier circuit (fcut=650, feed=95).
1933 Cut frequency (in Hz).
1942 Remap input channels to new locations.
1944 It accepts the following parameters:
1947 Map channels from input to output. The argument is a '|'-separated list of
1948 mappings, each in the @code{@var{in_channel}-@var{out_channel}} or
1949 @var{in_channel} form. @var{in_channel} can be either the name of the input
1950 channel (e.g. FL for front left) or its index in the input channel layout.
1951 @var{out_channel} is the name of the output channel or its index in the output
1952 channel layout. If @var{out_channel} is not given then it is implicitly an
1953 index, starting with zero and increasing by one for each mapping.
1955 @item channel_layout
1956 The channel layout of the output stream.
1959 If no mapping is present, the filter will implicitly map input channels to
1960 output channels, preserving indices.
1962 For example, assuming a 5.1+downmix input MOV file,
1964 ffmpeg -i in.mov -filter 'channelmap=map=DL-FL|DR-FR' out.wav
1966 will create an output WAV file tagged as stereo from the downmix channels of
1969 To fix a 5.1 WAV improperly encoded in AAC's native channel order
1971 ffmpeg -i in.wav -filter 'channelmap=1|2|0|5|3|4:5.1' out.wav
1974 @section channelsplit
1976 Split each channel from an input audio stream into a separate output stream.
1978 It accepts the following parameters:
1980 @item channel_layout
1981 The channel layout of the input stream. The default is "stereo".
1984 For example, assuming a stereo input MP3 file,
1986 ffmpeg -i in.mp3 -filter_complex channelsplit out.mkv
1988 will create an output Matroska file with two audio streams, one containing only
1989 the left channel and the other the right channel.
1991 Split a 5.1 WAV file into per-channel files:
1993 ffmpeg -i in.wav -filter_complex
1994 'channelsplit=channel_layout=5.1[FL][FR][FC][LFE][SL][SR]'
1995 -map '[FL]' front_left.wav -map '[FR]' front_right.wav -map '[FC]'
1996 front_center.wav -map '[LFE]' lfe.wav -map '[SL]' side_left.wav -map '[SR]'
2001 Add a chorus effect to the audio.
2003 Can make a single vocal sound like a chorus, but can also be applied to instrumentation.
2005 Chorus resembles an echo effect with a short delay, but whereas with echo the delay is
2006 constant, with chorus, it is varied using using sinusoidal or triangular modulation.
2007 The modulation depth defines the range the modulated delay is played before or after
2008 the delay. Hence the delayed sound will sound slower or faster, that is the delayed
2009 sound tuned around the original one, like in a chorus where some vocals are slightly
2012 It accepts the following parameters:
2015 Set input gain. Default is 0.4.
2018 Set output gain. Default is 0.4.
2021 Set delays. A typical delay is around 40ms to 60ms.
2033 @subsection Examples
2039 chorus=0.7:0.9:55:0.4:0.25:2
2045 chorus=0.6:0.9:50|60:0.4|0.32:0.25|0.4:2|1.3
2049 Fuller sounding chorus with three delays:
2051 chorus=0.5:0.9:50|60|40:0.4|0.32|0.3:0.25|0.4|0.3:2|2.3|1.3
2056 Compress or expand the audio's dynamic range.
2058 It accepts the following parameters:
2064 A list of times in seconds for each channel over which the instantaneous level
2065 of the input signal is averaged to determine its volume. @var{attacks} refers to
2066 increase of volume and @var{decays} refers to decrease of volume. For most
2067 situations, the attack time (response to the audio getting louder) should be
2068 shorter than the decay time, because the human ear is more sensitive to sudden
2069 loud audio than sudden soft audio. A typical value for attack is 0.3 seconds and
2070 a typical value for decay is 0.8 seconds.
2071 If specified number of attacks & decays is lower than number of channels, the last
2072 set attack/decay will be used for all remaining channels.
2075 A list of points for the transfer function, specified in dB relative to the
2076 maximum possible signal amplitude. Each key points list must be defined using
2077 the following syntax: @code{x0/y0|x1/y1|x2/y2|....} or
2078 @code{x0/y0 x1/y1 x2/y2 ....}
2080 The input values must be in strictly increasing order but the transfer function
2081 does not have to be monotonically rising. The point @code{0/0} is assumed but
2082 may be overridden (by @code{0/out-dBn}). Typical values for the transfer
2083 function are @code{-70/-70|-60/-20}.
2086 Set the curve radius in dB for all joints. It defaults to 0.01.
2089 Set the additional gain in dB to be applied at all points on the transfer
2090 function. This allows for easy adjustment of the overall gain.
2094 Set an initial volume, in dB, to be assumed for each channel when filtering
2095 starts. This permits the user to supply a nominal level initially, so that, for
2096 example, a very large gain is not applied to initial signal levels before the
2097 companding has begun to operate. A typical value for audio which is initially
2098 quiet is -90 dB. It defaults to 0.
2101 Set a delay, in seconds. The input audio is analyzed immediately, but audio is
2102 delayed before being fed to the volume adjuster. Specifying a delay
2103 approximately equal to the attack/decay times allows the filter to effectively
2104 operate in predictive rather than reactive mode. It defaults to 0.
2108 @subsection Examples
2112 Make music with both quiet and loud passages suitable for listening to in a
2115 compand=.3|.3:1|1:-90/-60|-60/-40|-40/-30|-20/-20:6:0:-90:0.2
2118 Another example for audio with whisper and explosion parts:
2120 compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0
2124 A noise gate for when the noise is at a lower level than the signal:
2126 compand=.1|.1:.2|.2:-900/-900|-50.1/-900|-50/-50:.01:0:-90:.1
2130 Here is another noise gate, this time for when the noise is at a higher level
2131 than the signal (making it, in some ways, similar to squelch):
2133 compand=.1|.1:.1|.1:-45.1/-45.1|-45/-900|0/-900:.01:45:-90:.1
2137 2:1 compression starting at -6dB:
2139 compand=points=-80/-80|-6/-6|0/-3.8|20/3.5
2143 2:1 compression starting at -9dB:
2145 compand=points=-80/-80|-9/-9|0/-5.3|20/2.9
2149 2:1 compression starting at -12dB:
2151 compand=points=-80/-80|-12/-12|0/-6.8|20/1.9
2155 2:1 compression starting at -18dB:
2157 compand=points=-80/-80|-18/-18|0/-9.8|20/0.7
2161 3:1 compression starting at -15dB:
2163 compand=points=-80/-80|-15/-15|0/-10.8|20/-5.2
2169 compand=points=-80/-105|-62/-80|-15.4/-15.4|0/-12|20/-7.6
2175 compand=attacks=0:points=-80/-169|-54/-80|-49.5/-64.6|-41.1/-41.1|-25.8/-15|-10.8/-4.5|0/0|20/8.3
2179 Hard limiter at -6dB:
2181 compand=attacks=0:points=-80/-80|-6/-6|20/-6
2185 Hard limiter at -12dB:
2187 compand=attacks=0:points=-80/-80|-12/-12|20/-12
2191 Hard noise gate at -35 dB:
2193 compand=attacks=0:points=-80/-115|-35.1/-80|-35/-35|20/20
2199 compand=attacks=0:points=-80/-80|-12.4/-12.4|-6/-8|0/-6.8|20/-2.8
2203 @section compensationdelay
2205 Compensation Delay Line is a metric based delay to compensate differing
2206 positions of microphones or speakers.
2208 For example, you have recorded guitar with two microphones placed in
2209 different location. Because the front of sound wave has fixed speed in
2210 normal conditions, the phasing of microphones can vary and depends on
2211 their location and interposition. The best sound mix can be achieved when
2212 these microphones are in phase (synchronized). Note that distance of
2213 ~30 cm between microphones makes one microphone to capture signal in
2214 antiphase to another microphone. That makes the final mix sounding moody.
2215 This filter helps to solve phasing problems by adding different delays
2216 to each microphone track and make them synchronized.
2218 The best result can be reached when you take one track as base and
2219 synchronize other tracks one by one with it.
2220 Remember that synchronization/delay tolerance depends on sample rate, too.
2221 Higher sample rates will give more tolerance.
2223 It accepts the following parameters:
2227 Set millimeters distance. This is compensation distance for fine tuning.
2231 Set cm distance. This is compensation distance for tightening distance setup.
2235 Set meters distance. This is compensation distance for hard distance setup.
2239 Set dry amount. Amount of unprocessed (dry) signal.
2243 Set wet amount. Amount of processed (wet) signal.
2247 Set temperature degree in Celsius. This is the temperature of the environment.
2251 @section crystalizer
2252 Simple algorithm to expand audio dynamic range.
2254 The filter accepts the following options:
2258 Sets the intensity of effect (default: 2.0). Must be in range between 0.0
2259 (unchanged sound) to 10.0 (maximum effect).
2262 Enable clipping. By default is enabled.
2266 Apply a DC shift to the audio.
2268 This can be useful to remove a DC offset (caused perhaps by a hardware problem
2269 in the recording chain) from the audio. The effect of a DC offset is reduced
2270 headroom and hence volume. The @ref{astats} filter can be used to determine if
2271 a signal has a DC offset.
2275 Set the DC shift, allowed range is [-1, 1]. It indicates the amount to shift
2279 Optional. It should have a value much less than 1 (e.g. 0.05 or 0.02) and is
2280 used to prevent clipping.
2284 Dynamic Audio Normalizer.
2286 This filter applies a certain amount of gain to the input audio in order
2287 to bring its peak magnitude to a target level (e.g. 0 dBFS). However, in
2288 contrast to more "simple" normalization algorithms, the Dynamic Audio
2289 Normalizer *dynamically* re-adjusts the gain factor to the input audio.
2290 This allows for applying extra gain to the "quiet" sections of the audio
2291 while avoiding distortions or clipping the "loud" sections. In other words:
2292 The Dynamic Audio Normalizer will "even out" the volume of quiet and loud
2293 sections, in the sense that the volume of each section is brought to the
2294 same target level. Note, however, that the Dynamic Audio Normalizer achieves
2295 this goal *without* applying "dynamic range compressing". It will retain 100%
2296 of the dynamic range *within* each section of the audio file.
2300 Set the frame length in milliseconds. In range from 10 to 8000 milliseconds.
2301 Default is 500 milliseconds.
2302 The Dynamic Audio Normalizer processes the input audio in small chunks,
2303 referred to as frames. This is required, because a peak magnitude has no
2304 meaning for just a single sample value. Instead, we need to determine the
2305 peak magnitude for a contiguous sequence of sample values. While a "standard"
2306 normalizer would simply use the peak magnitude of the complete file, the
2307 Dynamic Audio Normalizer determines the peak magnitude individually for each
2308 frame. The length of a frame is specified in milliseconds. By default, the
2309 Dynamic Audio Normalizer uses a frame length of 500 milliseconds, which has
2310 been found to give good results with most files.
2311 Note that the exact frame length, in number of samples, will be determined
2312 automatically, based on the sampling rate of the individual input audio file.
2315 Set the Gaussian filter window size. In range from 3 to 301, must be odd
2316 number. Default is 31.
2317 Probably the most important parameter of the Dynamic Audio Normalizer is the
2318 @code{window size} of the Gaussian smoothing filter. The filter's window size
2319 is specified in frames, centered around the current frame. For the sake of
2320 simplicity, this must be an odd number. Consequently, the default value of 31
2321 takes into account the current frame, as well as the 15 preceding frames and
2322 the 15 subsequent frames. Using a larger window results in a stronger
2323 smoothing effect and thus in less gain variation, i.e. slower gain
2324 adaptation. Conversely, using a smaller window results in a weaker smoothing
2325 effect and thus in more gain variation, i.e. faster gain adaptation.
2326 In other words, the more you increase this value, the more the Dynamic Audio
2327 Normalizer will behave like a "traditional" normalization filter. On the
2328 contrary, the more you decrease this value, the more the Dynamic Audio
2329 Normalizer will behave like a dynamic range compressor.
2332 Set the target peak value. This specifies the highest permissible magnitude
2333 level for the normalized audio input. This filter will try to approach the
2334 target peak magnitude as closely as possible, but at the same time it also
2335 makes sure that the normalized signal will never exceed the peak magnitude.
2336 A frame's maximum local gain factor is imposed directly by the target peak
2337 magnitude. The default value is 0.95 and thus leaves a headroom of 5%*.
2338 It is not recommended to go above this value.
2341 Set the maximum gain factor. In range from 1.0 to 100.0. Default is 10.0.
2342 The Dynamic Audio Normalizer determines the maximum possible (local) gain
2343 factor for each input frame, i.e. the maximum gain factor that does not
2344 result in clipping or distortion. The maximum gain factor is determined by
2345 the frame's highest magnitude sample. However, the Dynamic Audio Normalizer
2346 additionally bounds the frame's maximum gain factor by a predetermined
2347 (global) maximum gain factor. This is done in order to avoid excessive gain
2348 factors in "silent" or almost silent frames. By default, the maximum gain
2349 factor is 10.0, For most inputs the default value should be sufficient and
2350 it usually is not recommended to increase this value. Though, for input
2351 with an extremely low overall volume level, it may be necessary to allow even
2352 higher gain factors. Note, however, that the Dynamic Audio Normalizer does
2353 not simply apply a "hard" threshold (i.e. cut off values above the threshold).
2354 Instead, a "sigmoid" threshold function will be applied. This way, the
2355 gain factors will smoothly approach the threshold value, but never exceed that
2359 Set the target RMS. In range from 0.0 to 1.0. Default is 0.0 - disabled.
2360 By default, the Dynamic Audio Normalizer performs "peak" normalization.
2361 This means that the maximum local gain factor for each frame is defined
2362 (only) by the frame's highest magnitude sample. This way, the samples can
2363 be amplified as much as possible without exceeding the maximum signal
2364 level, i.e. without clipping. Optionally, however, the Dynamic Audio
2365 Normalizer can also take into account the frame's root mean square,
2366 abbreviated RMS. In electrical engineering, the RMS is commonly used to
2367 determine the power of a time-varying signal. It is therefore considered
2368 that the RMS is a better approximation of the "perceived loudness" than
2369 just looking at the signal's peak magnitude. Consequently, by adjusting all
2370 frames to a constant RMS value, a uniform "perceived loudness" can be
2371 established. If a target RMS value has been specified, a frame's local gain
2372 factor is defined as the factor that would result in exactly that RMS value.
2373 Note, however, that the maximum local gain factor is still restricted by the
2374 frame's highest magnitude sample, in order to prevent clipping.
2377 Enable channels coupling. By default is enabled.
2378 By default, the Dynamic Audio Normalizer will amplify all channels by the same
2379 amount. This means the same gain factor will be applied to all channels, i.e.
2380 the maximum possible gain factor is determined by the "loudest" channel.
2381 However, in some recordings, it may happen that the volume of the different
2382 channels is uneven, e.g. one channel may be "quieter" than the other one(s).
2383 In this case, this option can be used to disable the channel coupling. This way,
2384 the gain factor will be determined independently for each channel, depending
2385 only on the individual channel's highest magnitude sample. This allows for
2386 harmonizing the volume of the different channels.
2389 Enable DC bias correction. By default is disabled.
2390 An audio signal (in the time domain) is a sequence of sample values.
2391 In the Dynamic Audio Normalizer these sample values are represented in the
2392 -1.0 to 1.0 range, regardless of the original input format. Normally, the
2393 audio signal, or "waveform", should be centered around the zero point.
2394 That means if we calculate the mean value of all samples in a file, or in a
2395 single frame, then the result should be 0.0 or at least very close to that
2396 value. If, however, there is a significant deviation of the mean value from
2397 0.0, in either positive or negative direction, this is referred to as a
2398 DC bias or DC offset. Since a DC bias is clearly undesirable, the Dynamic
2399 Audio Normalizer provides optional DC bias correction.
2400 With DC bias correction enabled, the Dynamic Audio Normalizer will determine
2401 the mean value, or "DC correction" offset, of each input frame and subtract
2402 that value from all of the frame's sample values which ensures those samples
2403 are centered around 0.0 again. Also, in order to avoid "gaps" at the frame
2404 boundaries, the DC correction offset values will be interpolated smoothly
2405 between neighbouring frames.
2408 Enable alternative boundary mode. By default is disabled.
2409 The Dynamic Audio Normalizer takes into account a certain neighbourhood
2410 around each frame. This includes the preceding frames as well as the
2411 subsequent frames. However, for the "boundary" frames, located at the very
2412 beginning and at the very end of the audio file, not all neighbouring
2413 frames are available. In particular, for the first few frames in the audio
2414 file, the preceding frames are not known. And, similarly, for the last few
2415 frames in the audio file, the subsequent frames are not known. Thus, the
2416 question arises which gain factors should be assumed for the missing frames
2417 in the "boundary" region. The Dynamic Audio Normalizer implements two modes
2418 to deal with this situation. The default boundary mode assumes a gain factor
2419 of exactly 1.0 for the missing frames, resulting in a smooth "fade in" and
2420 "fade out" at the beginning and at the end of the input, respectively.
2423 Set the compress factor. In range from 0.0 to 30.0. Default is 0.0.
2424 By default, the Dynamic Audio Normalizer does not apply "traditional"
2425 compression. This means that signal peaks will not be pruned and thus the
2426 full dynamic range will be retained within each local neighbourhood. However,
2427 in some cases it may be desirable to combine the Dynamic Audio Normalizer's
2428 normalization algorithm with a more "traditional" compression.
2429 For this purpose, the Dynamic Audio Normalizer provides an optional compression
2430 (thresholding) function. If (and only if) the compression feature is enabled,
2431 all input frames will be processed by a soft knee thresholding function prior
2432 to the actual normalization process. Put simply, the thresholding function is
2433 going to prune all samples whose magnitude exceeds a certain threshold value.
2434 However, the Dynamic Audio Normalizer does not simply apply a fixed threshold
2435 value. Instead, the threshold value will be adjusted for each individual
2437 In general, smaller parameters result in stronger compression, and vice versa.
2438 Values below 3.0 are not recommended, because audible distortion may appear.
2443 Make audio easier to listen to on headphones.
2445 This filter adds `cues' to 44.1kHz stereo (i.e. audio CD format) audio
2446 so that when listened to on headphones the stereo image is moved from
2447 inside your head (standard for headphones) to outside and in front of
2448 the listener (standard for speakers).
2454 Apply a two-pole peaking equalisation (EQ) filter. With this
2455 filter, the signal-level at and around a selected frequency can
2456 be increased or decreased, whilst (unlike bandpass and bandreject
2457 filters) that at all other frequencies is unchanged.
2459 In order to produce complex equalisation curves, this filter can
2460 be given several times, each with a different central frequency.
2462 The filter accepts the following options:
2466 Set the filter's central frequency in Hz.
2469 Set method to specify band-width of filter.
2482 Specify the band-width of a filter in width_type units.
2485 Set the required gain or attenuation in dB.
2486 Beware of clipping when using a positive gain.
2489 Specify which channels to filter, by default all available are filtered.
2492 @subsection Examples
2495 Attenuate 10 dB at 1000 Hz, with a bandwidth of 200 Hz:
2497 equalizer=f=1000:width_type=h:width=200:g=-10
2501 Apply 2 dB gain at 1000 Hz with Q 1 and attenuate 5 dB at 100 Hz with Q 2:
2503 equalizer=f=1000:width_type=q:width=1:g=2,equalizer=f=100:width_type=q:width=2:g=-5
2507 @section extrastereo
2509 Linearly increases the difference between left and right channels which
2510 adds some sort of "live" effect to playback.
2512 The filter accepts the following options:
2516 Sets the difference coefficient (default: 2.5). 0.0 means mono sound
2517 (average of both channels), with 1.0 sound will be unchanged, with
2518 -1.0 left and right channels will be swapped.
2521 Enable clipping. By default is enabled.
2524 @section firequalizer
2525 Apply FIR Equalization using arbitrary frequency response.
2527 The filter accepts the following option:
2531 Set gain curve equation (in dB). The expression can contain variables:
2534 the evaluated frequency
2538 channel number, set to 0 when multichannels evaluation is disabled
2540 channel id, see libavutil/channel_layout.h, set to the first channel id when
2541 multichannels evaluation is disabled
2545 channel_layout, see libavutil/channel_layout.h
2550 @item gain_interpolate(f)
2551 interpolate gain on frequency f based on gain_entry
2552 @item cubic_interpolate(f)
2553 same as gain_interpolate, but smoother
2555 This option is also available as command. Default is @code{gain_interpolate(f)}.
2558 Set gain entry for gain_interpolate function. The expression can
2562 store gain entry at frequency f with value g
2564 This option is also available as command.
2567 Set filter delay in seconds. Higher value means more accurate.
2568 Default is @code{0.01}.
2571 Set filter accuracy in Hz. Lower value means more accurate.
2572 Default is @code{5}.
2575 Set window function. Acceptable values are:
2578 rectangular window, useful when gain curve is already smooth
2580 hann window (default)
2586 3-terms continuous 1st derivative nuttall window
2588 minimum 3-terms discontinuous nuttall window
2590 4-terms continuous 1st derivative nuttall window
2592 minimum 4-terms discontinuous nuttall (blackman-nuttall) window
2594 blackman-harris window
2600 If enabled, use fixed number of audio samples. This improves speed when
2601 filtering with large delay. Default is disabled.
2604 Enable multichannels evaluation on gain. Default is disabled.
2607 Enable zero phase mode by subtracting timestamp to compensate delay.
2608 Default is disabled.
2611 Set scale used by gain. Acceptable values are:
2614 linear frequency, linear gain
2616 linear frequency, logarithmic (in dB) gain (default)
2618 logarithmic (in octave scale where 20 Hz is 0) frequency, linear gain
2620 logarithmic frequency, logarithmic gain
2624 Set file for dumping, suitable for gnuplot.
2627 Set scale for dumpfile. Acceptable values are same with scale option.
2631 Enable 2-channel convolution using complex FFT. This improves speed significantly.
2632 Default is disabled.
2635 @subsection Examples
2640 firequalizer=gain='if(lt(f,1000), 0, -INF)'
2643 lowpass at 1000 Hz with gain_entry:
2645 firequalizer=gain_entry='entry(1000,0); entry(1001, -INF)'
2648 custom equalization:
2650 firequalizer=gain_entry='entry(100,0); entry(400, -4); entry(1000, -6); entry(2000, 0)'
2653 higher delay with zero phase to compensate delay:
2655 firequalizer=delay=0.1:fixed=on:zero_phase=on
2658 lowpass on left channel, highpass on right channel:
2660 firequalizer=gain='if(eq(chid,1), gain_interpolate(f), if(eq(chid,2), gain_interpolate(1e6+f), 0))'
2661 :gain_entry='entry(1000, 0); entry(1001,-INF); entry(1e6+1000,0)':multi=on
2666 Apply a flanging effect to the audio.
2668 The filter accepts the following options:
2672 Set base delay in milliseconds. Range from 0 to 30. Default value is 0.
2675 Set added swep delay in milliseconds. Range from 0 to 10. Default value is 2.
2678 Set percentage regeneration (delayed signal feedback). Range from -95 to 95.
2682 Set percentage of delayed signal mixed with original. Range from 0 to 100.
2683 Default value is 71.
2686 Set sweeps per second (Hz). Range from 0.1 to 10. Default value is 0.5.
2689 Set swept wave shape, can be @var{triangular} or @var{sinusoidal}.
2690 Default value is @var{sinusoidal}.
2693 Set swept wave percentage-shift for multi channel. Range from 0 to 100.
2694 Default value is 25.
2697 Set delay-line interpolation, @var{linear} or @var{quadratic}.
2698 Default is @var{linear}.
2703 Decodes High Definition Compatible Digital (HDCD) data. A 16-bit PCM stream with
2704 embedded HDCD codes is expanded into a 20-bit PCM stream.
2706 The filter supports the Peak Extend and Low-level Gain Adjustment features
2707 of HDCD, and detects the Transient Filter flag.
2710 ffmpeg -i HDCD16.flac -af hdcd OUT24.flac
2713 When using the filter with wav, note the default encoding for wav is 16-bit,
2714 so the resulting 20-bit stream will be truncated back to 16-bit. Use something
2715 like @command{-acodec pcm_s24le} after the filter to get 24-bit PCM output.
2717 ffmpeg -i HDCD16.wav -af hdcd OUT16.wav
2718 ffmpeg -i HDCD16.wav -af hdcd -acodec pcm_s24le OUT24.wav
2721 The filter accepts the following options:
2724 @item disable_autoconvert
2725 Disable any automatic format conversion or resampling in the filter graph.
2727 @item process_stereo
2728 Process the stereo channels together. If target_gain does not match between
2729 channels, consider it invalid and use the last valid target_gain.
2732 Set the code detect timer period in ms.
2735 Always extend peaks above -3dBFS even if PE isn't signaled.
2738 Replace audio with a solid tone and adjust the amplitude to signal some
2739 specific aspect of the decoding process. The output file can be loaded in
2740 an audio editor alongside the original to aid analysis.
2742 @code{analyze_mode=pe:force_pe=true} can be used to see all samples above the PE level.
2749 Gain adjustment level at each sample
2751 Samples where peak extend occurs
2753 Samples where the code detect timer is active
2755 Samples where the target gain does not match between channels
2761 Apply a high-pass filter with 3dB point frequency.
2762 The filter can be either single-pole, or double-pole (the default).
2763 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
2765 The filter accepts the following options:
2769 Set frequency in Hz. Default is 3000.
2772 Set number of poles. Default is 2.
2775 Set method to specify band-width of filter.
2788 Specify the band-width of a filter in width_type units.
2789 Applies only to double-pole filter.
2790 The default is 0.707q and gives a Butterworth response.
2793 Specify which channels to filter, by default all available are filtered.
2798 Join multiple input streams into one multi-channel stream.
2800 It accepts the following parameters:
2804 The number of input streams. It defaults to 2.
2806 @item channel_layout
2807 The desired output channel layout. It defaults to stereo.
2810 Map channels from inputs to output. The argument is a '|'-separated list of
2811 mappings, each in the @code{@var{input_idx}.@var{in_channel}-@var{out_channel}}
2812 form. @var{input_idx} is the 0-based index of the input stream. @var{in_channel}
2813 can be either the name of the input channel (e.g. FL for front left) or its
2814 index in the specified input stream. @var{out_channel} is the name of the output
2818 The filter will attempt to guess the mappings when they are not specified
2819 explicitly. It does so by first trying to find an unused matching input channel
2820 and if that fails it picks the first unused input channel.
2822 Join 3 inputs (with properly set channel layouts):
2824 ffmpeg -i INPUT1 -i INPUT2 -i INPUT3 -filter_complex join=inputs=3 OUTPUT
2827 Build a 5.1 output from 6 single-channel streams:
2829 ffmpeg -i fl -i fr -i fc -i sl -i sr -i lfe -filter_complex
2830 'join=inputs=6:channel_layout=5.1:map=0.0-FL|1.0-FR|2.0-FC|3.0-SL|4.0-SR|5.0-LFE'
2836 Load a LADSPA (Linux Audio Developer's Simple Plugin API) plugin.
2838 To enable compilation of this filter you need to configure FFmpeg with
2839 @code{--enable-ladspa}.
2843 Specifies the name of LADSPA plugin library to load. If the environment
2844 variable @env{LADSPA_PATH} is defined, the LADSPA plugin is searched in
2845 each one of the directories specified by the colon separated list in
2846 @env{LADSPA_PATH}, otherwise in the standard LADSPA paths, which are in
2847 this order: @file{HOME/.ladspa/lib/}, @file{/usr/local/lib/ladspa/},
2848 @file{/usr/lib/ladspa/}.
2851 Specifies the plugin within the library. Some libraries contain only
2852 one plugin, but others contain many of them. If this is not set filter
2853 will list all available plugins within the specified library.
2856 Set the '|' separated list of controls which are zero or more floating point
2857 values that determine the behavior of the loaded plugin (for example delay,
2859 Controls need to be defined using the following syntax:
2860 c0=@var{value0}|c1=@var{value1}|c2=@var{value2}|..., where
2861 @var{valuei} is the value set on the @var{i}-th control.
2862 Alternatively they can be also defined using the following syntax:
2863 @var{value0}|@var{value1}|@var{value2}|..., where
2864 @var{valuei} is the value set on the @var{i}-th control.
2865 If @option{controls} is set to @code{help}, all available controls and
2866 their valid ranges are printed.
2868 @item sample_rate, s
2869 Specify the sample rate, default to 44100. Only used if plugin have
2873 Set the number of samples per channel per each output frame, default
2874 is 1024. Only used if plugin have zero inputs.
2877 Set the minimum duration of the sourced audio. See
2878 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
2879 for the accepted syntax.
2880 Note that the resulting duration may be greater than the specified duration,
2881 as the generated audio is always cut at the end of a complete frame.
2882 If not specified, or the expressed duration is negative, the audio is
2883 supposed to be generated forever.
2884 Only used if plugin have zero inputs.
2888 @subsection Examples
2892 List all available plugins within amp (LADSPA example plugin) library:
2898 List all available controls and their valid ranges for @code{vcf_notch}
2899 plugin from @code{VCF} library:
2901 ladspa=f=vcf:p=vcf_notch:c=help
2905 Simulate low quality audio equipment using @code{Computer Music Toolkit} (CMT)
2908 ladspa=file=cmt:plugin=lofi:controls=c0=22|c1=12|c2=12
2912 Add reverberation to the audio using TAP-plugins
2913 (Tom's Audio Processing plugins):
2915 ladspa=file=tap_reverb:tap_reverb
2919 Generate white noise, with 0.2 amplitude:
2921 ladspa=file=cmt:noise_source_white:c=c0=.2
2925 Generate 20 bpm clicks using plugin @code{C* Click - Metronome} from the
2926 @code{C* Audio Plugin Suite} (CAPS) library:
2928 ladspa=file=caps:Click:c=c1=20'
2932 Apply @code{C* Eq10X2 - Stereo 10-band equaliser} effect:
2934 ladspa=caps:Eq10X2:c=c0=-48|c9=-24|c3=12|c4=2
2938 Increase volume by 20dB using fast lookahead limiter from Steve Harris
2939 @code{SWH Plugins} collection:
2941 ladspa=fast_lookahead_limiter_1913:fastLookaheadLimiter:20|0|2
2945 Attenuate low frequencies using Multiband EQ from Steve Harris
2946 @code{SWH Plugins} collection:
2948 ladspa=mbeq_1197:mbeq:-24|-24|-24|0|0|0|0|0|0|0|0|0|0|0|0
2952 @subsection Commands
2954 This filter supports the following commands:
2957 Modify the @var{N}-th control value.
2959 If the specified value is not valid, it is ignored and prior one is kept.
2964 EBU R128 loudness normalization. Includes both dynamic and linear normalization modes.
2965 Support for both single pass (livestreams, files) and double pass (files) modes.
2966 This algorithm can target IL, LRA, and maximum true peak.
2968 The filter accepts the following options:
2972 Set integrated loudness target.
2973 Range is -70.0 - -5.0. Default value is -24.0.
2976 Set loudness range target.
2977 Range is 1.0 - 20.0. Default value is 7.0.
2980 Set maximum true peak.
2981 Range is -9.0 - +0.0. Default value is -2.0.
2983 @item measured_I, measured_i
2984 Measured IL of input file.
2985 Range is -99.0 - +0.0.
2987 @item measured_LRA, measured_lra
2988 Measured LRA of input file.
2989 Range is 0.0 - 99.0.
2991 @item measured_TP, measured_tp
2992 Measured true peak of input file.
2993 Range is -99.0 - +99.0.
2995 @item measured_thresh
2996 Measured threshold of input file.
2997 Range is -99.0 - +0.0.
3000 Set offset gain. Gain is applied before the true-peak limiter.
3001 Range is -99.0 - +99.0. Default is +0.0.
3004 Normalize linearly if possible.
3005 measured_I, measured_LRA, measured_TP, and measured_thresh must also
3006 to be specified in order to use this mode.
3007 Options are true or false. Default is true.
3010 Treat mono input files as "dual-mono". If a mono file is intended for playback
3011 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
3012 If set to @code{true}, this option will compensate for this effect.
3013 Multi-channel input files are not affected by this option.
3014 Options are true or false. Default is false.
3017 Set print format for stats. Options are summary, json, or none.
3018 Default value is none.
3023 Apply a low-pass filter with 3dB point frequency.
3024 The filter can be either single-pole or double-pole (the default).
3025 The filter roll off at 6dB per pole per octave (20dB per pole per decade).
3027 The filter accepts the following options:
3031 Set frequency in Hz. Default is 500.
3034 Set number of poles. Default is 2.
3037 Set method to specify band-width of filter.
3050 Specify the band-width of a filter in width_type units.
3051 Applies only to double-pole filter.
3052 The default is 0.707q and gives a Butterworth response.
3055 Specify which channels to filter, by default all available are filtered.
3058 @subsection Examples
3061 Lowpass only LFE channel, it LFE is not present it does nothing:
3070 Mix channels with specific gain levels. The filter accepts the output
3071 channel layout followed by a set of channels definitions.
3073 This filter is also designed to efficiently remap the channels of an audio
3076 The filter accepts parameters of the form:
3077 "@var{l}|@var{outdef}|@var{outdef}|..."
3081 output channel layout or number of channels
3084 output channel specification, of the form:
3085 "@var{out_name}=[@var{gain}*]@var{in_name}[(+-)[@var{gain}*]@var{in_name}...]"
3088 output channel to define, either a channel name (FL, FR, etc.) or a channel
3089 number (c0, c1, etc.)
3092 multiplicative coefficient for the channel, 1 leaving the volume unchanged
3095 input channel to use, see out_name for details; it is not possible to mix
3096 named and numbered input channels
3099 If the `=' in a channel specification is replaced by `<', then the gains for
3100 that specification will be renormalized so that the total is 1, thus
3101 avoiding clipping noise.
3103 @subsection Mixing examples
3105 For example, if you want to down-mix from stereo to mono, but with a bigger
3106 factor for the left channel:
3108 pan=1c|c0=0.9*c0+0.1*c1
3111 A customized down-mix to stereo that works automatically for 3-, 4-, 5- and
3112 7-channels surround:
3114 pan=stereo| FL < FL + 0.5*FC + 0.6*BL + 0.6*SL | FR < FR + 0.5*FC + 0.6*BR + 0.6*SR
3117 Note that @command{ffmpeg} integrates a default down-mix (and up-mix) system
3118 that should be preferred (see "-ac" option) unless you have very specific
3121 @subsection Remapping examples
3123 The channel remapping will be effective if, and only if:
3126 @item gain coefficients are zeroes or ones,
3127 @item only one input per channel output,
3130 If all these conditions are satisfied, the filter will notify the user ("Pure
3131 channel mapping detected"), and use an optimized and lossless method to do the
3134 For example, if you have a 5.1 source and want a stereo audio stream by
3135 dropping the extra channels:
3137 pan="stereo| c0=FL | c1=FR"
3140 Given the same source, you can also switch front left and front right channels
3141 and keep the input channel layout:
3143 pan="5.1| c0=c1 | c1=c0 | c2=c2 | c3=c3 | c4=c4 | c5=c5"
3146 If the input is a stereo audio stream, you can mute the front left channel (and
3147 still keep the stereo channel layout) with:
3152 Still with a stereo audio stream input, you can copy the right channel in both
3153 front left and right:
3155 pan="stereo| c0=FR | c1=FR"
3160 ReplayGain scanner filter. This filter takes an audio stream as an input and
3161 outputs it unchanged.
3162 At end of filtering it displays @code{track_gain} and @code{track_peak}.
3166 Convert the audio sample format, sample rate and channel layout. It is
3167 not meant to be used directly.
3170 Apply time-stretching and pitch-shifting with librubberband.
3172 The filter accepts the following options:
3176 Set tempo scale factor.
3179 Set pitch scale factor.
3182 Set transients detector.
3183 Possible values are:
3192 Possible values are:
3201 Possible values are:
3208 Set processing window size.
3209 Possible values are:
3218 Possible values are:
3225 Enable formant preservation when shift pitching.
3226 Possible values are:
3234 Possible values are:
3243 Possible values are:
3250 @section sidechaincompress
3252 This filter acts like normal compressor but has the ability to compress
3253 detected signal using second input signal.
3254 It needs two input streams and returns one output stream.
3255 First input stream will be processed depending on second stream signal.
3256 The filtered signal then can be filtered with other filters in later stages of
3257 processing. See @ref{pan} and @ref{amerge} filter.
3259 The filter accepts the following options:
3263 Set input gain. Default is 1. Range is between 0.015625 and 64.
3266 If a signal of second stream raises above this level it will affect the gain
3267 reduction of first stream.
3268 By default is 0.125. Range is between 0.00097563 and 1.
3271 Set a ratio about which the signal is reduced. 1:2 means that if the level
3272 raised 4dB above the threshold, it will be only 2dB above after the reduction.
3273 Default is 2. Range is between 1 and 20.
3276 Amount of milliseconds the signal has to rise above the threshold before gain
3277 reduction starts. Default is 20. Range is between 0.01 and 2000.
3280 Amount of milliseconds the signal has to fall below the threshold before
3281 reduction is decreased again. Default is 250. Range is between 0.01 and 9000.
3284 Set the amount by how much signal will be amplified after processing.
3285 Default is 2. Range is from 1 and 64.
3288 Curve the sharp knee around the threshold to enter gain reduction more softly.
3289 Default is 2.82843. Range is between 1 and 8.
3292 Choose if the @code{average} level between all channels of side-chain stream
3293 or the louder(@code{maximum}) channel of side-chain stream affects the
3294 reduction. Default is @code{average}.
3297 Should the exact signal be taken in case of @code{peak} or an RMS one in case
3298 of @code{rms}. Default is @code{rms} which is mainly smoother.
3301 Set sidechain gain. Default is 1. Range is between 0.015625 and 64.
3304 How much to use compressed signal in output. Default is 1.
3305 Range is between 0 and 1.
3308 @subsection Examples
3312 Full ffmpeg example taking 2 audio inputs, 1st input to be compressed
3313 depending on the signal of 2nd input and later compressed signal to be
3314 merged with 2nd input:
3316 ffmpeg -i main.flac -i sidechain.flac -filter_complex "[1:a]asplit=2[sc][mix];[0:a][sc]sidechaincompress[compr];[compr][mix]amerge"
3320 @section sidechaingate
3322 A sidechain gate acts like a normal (wideband) gate but has the ability to
3323 filter the detected signal before sending it to the gain reduction stage.
3324 Normally a gate uses the full range signal to detect a level above the
3326 For example: If you cut all lower frequencies from your sidechain signal
3327 the gate will decrease the volume of your track only if not enough highs
3328 appear. With this technique you are able to reduce the resonation of a
3329 natural drum or remove "rumbling" of muted strokes from a heavily distorted
3331 It needs two input streams and returns one output stream.
3332 First input stream will be processed depending on second stream signal.
3334 The filter accepts the following options:
3338 Set input level before filtering.
3339 Default is 1. Allowed range is from 0.015625 to 64.
3342 Set the level of gain reduction when the signal is below the threshold.
3343 Default is 0.06125. Allowed range is from 0 to 1.
3346 If a signal rises above this level the gain reduction is released.
3347 Default is 0.125. Allowed range is from 0 to 1.
3350 Set a ratio about which the signal is reduced.
3351 Default is 2. Allowed range is from 1 to 9000.
3354 Amount of milliseconds the signal has to rise above the threshold before gain
3356 Default is 20 milliseconds. Allowed range is from 0.01 to 9000.
3359 Amount of milliseconds the signal has to fall below the threshold before the
3360 reduction is increased again. Default is 250 milliseconds.
3361 Allowed range is from 0.01 to 9000.
3364 Set amount of amplification of signal after processing.
3365 Default is 1. Allowed range is from 1 to 64.
3368 Curve the sharp knee around the threshold to enter gain reduction more softly.
3369 Default is 2.828427125. Allowed range is from 1 to 8.
3372 Choose if exact signal should be taken for detection or an RMS like one.
3373 Default is rms. Can be peak or rms.
3376 Choose if the average level between all channels or the louder channel affects
3378 Default is average. Can be average or maximum.
3381 Set sidechain gain. Default is 1. Range is from 0.015625 to 64.
3384 @section silencedetect
3386 Detect silence in an audio stream.
3388 This filter logs a message when it detects that the input audio volume is less
3389 or equal to a noise tolerance value for a duration greater or equal to the
3390 minimum detected noise duration.
3392 The printed times and duration are expressed in seconds.
3394 The filter accepts the following options:
3398 Set silence duration until notification (default is 2 seconds).
3401 Set noise tolerance. Can be specified in dB (in case "dB" is appended to the
3402 specified value) or amplitude ratio. Default is -60dB, or 0.001.
3405 @subsection Examples
3409 Detect 5 seconds of silence with -50dB noise tolerance:
3411 silencedetect=n=-50dB:d=5
3415 Complete example with @command{ffmpeg} to detect silence with 0.0001 noise
3416 tolerance in @file{silence.mp3}:
3418 ffmpeg -i silence.mp3 -af silencedetect=noise=0.0001 -f null -
3422 @section silenceremove
3424 Remove silence from the beginning, middle or end of the audio.
3426 The filter accepts the following options:
3430 This value is used to indicate if audio should be trimmed at beginning of
3431 the audio. A value of zero indicates no silence should be trimmed from the
3432 beginning. When specifying a non-zero value, it trims audio up until it
3433 finds non-silence. Normally, when trimming silence from beginning of audio
3434 the @var{start_periods} will be @code{1} but it can be increased to higher
3435 values to trim all audio up to specific count of non-silence periods.
3436 Default value is @code{0}.
3438 @item start_duration
3439 Specify the amount of time that non-silence must be detected before it stops
3440 trimming audio. By increasing the duration, bursts of noises can be treated
3441 as silence and trimmed off. Default value is @code{0}.
3443 @item start_threshold
3444 This indicates what sample value should be treated as silence. For digital
3445 audio, a value of @code{0} may be fine but for audio recorded from analog,
3446 you may wish to increase the value to account for background noise.
3447 Can be specified in dB (in case "dB" is appended to the specified value)
3448 or amplitude ratio. Default value is @code{0}.
3451 Set the count for trimming silence from the end of audio.
3452 To remove silence from the middle of a file, specify a @var{stop_periods}
3453 that is negative. This value is then treated as a positive value and is
3454 used to indicate the effect should restart processing as specified by
3455 @var{start_periods}, making it suitable for removing periods of silence
3456 in the middle of the audio.
3457 Default value is @code{0}.
3460 Specify a duration of silence that must exist before audio is not copied any
3461 more. By specifying a higher duration, silence that is wanted can be left in
3463 Default value is @code{0}.
3465 @item stop_threshold
3466 This is the same as @option{start_threshold} but for trimming silence from
3468 Can be specified in dB (in case "dB" is appended to the specified value)
3469 or amplitude ratio. Default value is @code{0}.
3472 This indicates that @var{stop_duration} length of audio should be left intact
3473 at the beginning of each period of silence.
3474 For example, if you want to remove long pauses between words but do not want
3475 to remove the pauses completely. Default value is @code{0}.
3478 Set how is silence detected. Can be @code{rms} or @code{peak}. Second is faster
3479 and works better with digital silence which is exactly 0.
3480 Default value is @code{rms}.
3483 Set ratio used to calculate size of window for detecting silence.
3484 Default value is @code{0.02}. Allowed range is from @code{0} to @code{10}.
3487 @subsection Examples
3491 The following example shows how this filter can be used to start a recording
3492 that does not contain the delay at the start which usually occurs between
3493 pressing the record button and the start of the performance:
3495 silenceremove=1:5:0.02
3499 Trim all silence encountered from beginning to end where there is more than 1
3500 second of silence in audio:
3502 silenceremove=0:0:0:-1:1:-90dB
3508 SOFAlizer uses head-related transfer functions (HRTFs) to create virtual
3509 loudspeakers around the user for binaural listening via headphones (audio
3510 formats up to 9 channels supported).
3511 The HRTFs are stored in SOFA files (see @url{http://www.sofacoustics.org/} for a database).
3512 SOFAlizer is developed at the Acoustics Research Institute (ARI) of the
3513 Austrian Academy of Sciences.
3515 To enable compilation of this filter you need to configure FFmpeg with
3516 @code{--enable-netcdf}.
3518 The filter accepts the following options:
3522 Set the SOFA file used for rendering.
3525 Set gain applied to audio. Value is in dB. Default is 0.
3528 Set rotation of virtual loudspeakers in deg. Default is 0.
3531 Set elevation of virtual speakers in deg. Default is 0.
3534 Set distance in meters between loudspeakers and the listener with near-field
3535 HRTFs. Default is 1.
3538 Set processing type. Can be @var{time} or @var{freq}. @var{time} is
3539 processing audio in time domain which is slow.
3540 @var{freq} is processing audio in frequency domain which is fast.
3541 Default is @var{freq}.
3544 Set custom positions of virtual loudspeakers. Syntax for this option is:
3545 <CH> <AZIM> <ELEV>[|<CH> <AZIM> <ELEV>|...].
3546 Each virtual loudspeaker is described with short channel name following with
3547 azimuth and elevation in degreees.
3548 Each virtual loudspeaker description is separated by '|'.
3549 For example to override front left and front right channel positions use:
3550 'speakers=FL 45 15|FR 345 15'.
3551 Descriptions with unrecognised channel names are ignored.
3554 @subsection Examples
3558 Using ClubFritz6 sofa file:
3560 sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=1
3564 Using ClubFritz12 sofa file and bigger radius with small rotation:
3566 sofalizer=sofa=/path/to/ClubFritz12.sofa:type=freq:radius=2:rotation=5
3570 Similar as above but with custom speaker positions for front left, front right, back left and back right
3571 and also with custom gain:
3573 "sofalizer=sofa=/path/to/ClubFritz6.sofa:type=freq:radius=2:speakers=FL 45|FR 315|BL 135|BR 225:gain=28"
3577 @section stereotools
3579 This filter has some handy utilities to manage stereo signals, for converting
3580 M/S stereo recordings to L/R signal while having control over the parameters
3581 or spreading the stereo image of master track.
3583 The filter accepts the following options:
3587 Set input level before filtering for both channels. Defaults is 1.
3588 Allowed range is from 0.015625 to 64.
3591 Set output level after filtering for both channels. Defaults is 1.
3592 Allowed range is from 0.015625 to 64.
3595 Set input balance between both channels. Default is 0.
3596 Allowed range is from -1 to 1.
3599 Set output balance between both channels. Default is 0.
3600 Allowed range is from -1 to 1.
3603 Enable softclipping. Results in analog distortion instead of harsh digital 0dB
3604 clipping. Disabled by default.
3607 Mute the left channel. Disabled by default.
3610 Mute the right channel. Disabled by default.
3613 Change the phase of the left channel. Disabled by default.
3616 Change the phase of the right channel. Disabled by default.
3619 Set stereo mode. Available values are:
3623 Left/Right to Left/Right, this is default.
3626 Left/Right to Mid/Side.
3629 Mid/Side to Left/Right.
3632 Left/Right to Left/Left.
3635 Left/Right to Right/Right.
3638 Left/Right to Left + Right.
3641 Left/Right to Right/Left.
3645 Set level of side signal. Default is 1.
3646 Allowed range is from 0.015625 to 64.
3649 Set balance of side signal. Default is 0.
3650 Allowed range is from -1 to 1.
3653 Set level of the middle signal. Default is 1.
3654 Allowed range is from 0.015625 to 64.
3657 Set middle signal pan. Default is 0. Allowed range is from -1 to 1.
3660 Set stereo base between mono and inversed channels. Default is 0.
3661 Allowed range is from -1 to 1.
3664 Set delay in milliseconds how much to delay left from right channel and
3665 vice versa. Default is 0. Allowed range is from -20 to 20.
3668 Set S/C level. Default is 1. Allowed range is from 1 to 100.
3671 Set the stereo phase in degrees. Default is 0. Allowed range is from 0 to 360.
3674 @subsection Examples
3678 Apply karaoke like effect:
3680 stereotools=mlev=0.015625
3684 Convert M/S signal to L/R:
3686 "stereotools=mode=ms>lr"
3690 @section stereowiden
3692 This filter enhance the stereo effect by suppressing signal common to both
3693 channels and by delaying the signal of left into right and vice versa,
3694 thereby widening the stereo effect.
3696 The filter accepts the following options:
3700 Time in milliseconds of the delay of left signal into right and vice versa.
3701 Default is 20 milliseconds.
3704 Amount of gain in delayed signal into right and vice versa. Gives a delay
3705 effect of left signal in right output and vice versa which gives widening
3706 effect. Default is 0.3.
3709 Cross feed of left into right with inverted phase. This helps in suppressing
3710 the mono. If the value is 1 it will cancel all the signal common to both
3711 channels. Default is 0.3.
3714 Set level of input signal of original channel. Default is 0.8.
3719 Boost or cut treble (upper) frequencies of the audio using a two-pole
3720 shelving filter with a response similar to that of a standard
3721 hi-fi's tone-controls. This is also known as shelving equalisation (EQ).
3723 The filter accepts the following options:
3727 Give the gain at whichever is the lower of ~22 kHz and the
3728 Nyquist frequency. Its useful range is about -20 (for a large cut)
3729 to +20 (for a large boost). Beware of clipping when using a positive gain.
3732 Set the filter's central frequency and so can be used
3733 to extend or reduce the frequency range to be boosted or cut.
3734 The default value is @code{3000} Hz.
3737 Set method to specify band-width of filter.
3750 Determine how steep is the filter's shelf transition.
3753 Specify which channels to filter, by default all available are filtered.
3758 Sinusoidal amplitude modulation.
3760 The filter accepts the following options:
3764 Modulation frequency in Hertz. Modulation frequencies in the subharmonic range
3765 (20 Hz or lower) will result in a tremolo effect.
3766 This filter may also be used as a ring modulator by specifying
3767 a modulation frequency higher than 20 Hz.
3768 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3771 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3772 Default value is 0.5.
3777 Sinusoidal phase modulation.
3779 The filter accepts the following options:
3783 Modulation frequency in Hertz.
3784 Range is 0.1 - 20000.0. Default value is 5.0 Hz.
3787 Depth of modulation as a percentage. Range is 0.0 - 1.0.
3788 Default value is 0.5.
3793 Adjust the input audio volume.
3795 It accepts the following parameters:
3799 Set audio volume expression.
3801 Output values are clipped to the maximum value.
3803 The output audio volume is given by the relation:
3805 @var{output_volume} = @var{volume} * @var{input_volume}
3808 The default value for @var{volume} is "1.0".
3811 This parameter represents the mathematical precision.
3813 It determines which input sample formats will be allowed, which affects the
3814 precision of the volume scaling.
3818 8-bit fixed-point; this limits input sample format to U8, S16, and S32.
3820 32-bit floating-point; this limits input sample format to FLT. (default)
3822 64-bit floating-point; this limits input sample format to DBL.
3826 Choose the behaviour on encountering ReplayGain side data in input frames.
3830 Remove ReplayGain side data, ignoring its contents (the default).
3833 Ignore ReplayGain side data, but leave it in the frame.
3836 Prefer the track gain, if present.
3839 Prefer the album gain, if present.
3842 @item replaygain_preamp
3843 Pre-amplification gain in dB to apply to the selected replaygain gain.
3845 Default value for @var{replaygain_preamp} is 0.0.
3848 Set when the volume expression is evaluated.
3850 It accepts the following values:
3853 only evaluate expression once during the filter initialization, or
3854 when the @samp{volume} command is sent
3857 evaluate expression for each incoming frame
3860 Default value is @samp{once}.
3863 The volume expression can contain the following parameters.
3867 frame number (starting at zero)
3870 @item nb_consumed_samples
3871 number of samples consumed by the filter
3873 number of samples in the current frame
3875 original frame position in the file
3881 PTS at start of stream
3883 time at start of stream
3889 last set volume value
3892 Note that when @option{eval} is set to @samp{once} only the
3893 @var{sample_rate} and @var{tb} variables are available, all other
3894 variables will evaluate to NAN.
3896 @subsection Commands
3898 This filter supports the following commands:
3901 Modify the volume expression.
3902 The command accepts the same syntax of the corresponding option.
3904 If the specified expression is not valid, it is kept at its current
3906 @item replaygain_noclip
3907 Prevent clipping by limiting the gain applied.
3909 Default value for @var{replaygain_noclip} is 1.
3913 @subsection Examples
3917 Halve the input audio volume:
3921 volume=volume=-6.0206dB
3924 In all the above example the named key for @option{volume} can be
3925 omitted, for example like in:
3931 Increase input audio power by 6 decibels using fixed-point precision:
3933 volume=volume=6dB:precision=fixed
3937 Fade volume after time 10 with an annihilation period of 5 seconds:
3939 volume='if(lt(t,10),1,max(1-(t-10)/5,0))':eval=frame
3943 @section volumedetect
3945 Detect the volume of the input video.
3947 The filter has no parameters. The input is not modified. Statistics about
3948 the volume will be printed in the log when the input stream end is reached.
3950 In particular it will show the mean volume (root mean square), maximum
3951 volume (on a per-sample basis), and the beginning of a histogram of the
3952 registered volume values (from the maximum value to a cumulated 1/1000 of
3955 All volumes are in decibels relative to the maximum PCM value.
3957 @subsection Examples
3959 Here is an excerpt of the output:
3961 [Parsed_volumedetect_0 @ 0xa23120] mean_volume: -27 dB
3962 [Parsed_volumedetect_0 @ 0xa23120] max_volume: -4 dB
3963 [Parsed_volumedetect_0 @ 0xa23120] histogram_4db: 6
3964 [Parsed_volumedetect_0 @ 0xa23120] histogram_5db: 62
3965 [Parsed_volumedetect_0 @ 0xa23120] histogram_6db: 286
3966 [Parsed_volumedetect_0 @ 0xa23120] histogram_7db: 1042
3967 [Parsed_volumedetect_0 @ 0xa23120] histogram_8db: 2551
3968 [Parsed_volumedetect_0 @ 0xa23120] histogram_9db: 4609
3969 [Parsed_volumedetect_0 @ 0xa23120] histogram_10db: 8409
3975 The mean square energy is approximately -27 dB, or 10^-2.7.
3977 The largest sample is at -4 dB, or more precisely between -4 dB and -5 dB.
3979 There are 6 samples at -4 dB, 62 at -5 dB, 286 at -6 dB, etc.
3982 In other words, raising the volume by +4 dB does not cause any clipping,
3983 raising it by +5 dB causes clipping for 6 samples, etc.
3985 @c man end AUDIO FILTERS
3987 @chapter Audio Sources
3988 @c man begin AUDIO SOURCES
3990 Below is a description of the currently available audio sources.
3994 Buffer audio frames, and make them available to the filter chain.
3996 This source is mainly intended for a programmatic use, in particular
3997 through the interface defined in @file{libavfilter/asrc_abuffer.h}.
3999 It accepts the following parameters:
4003 The timebase which will be used for timestamps of submitted frames. It must be
4004 either a floating-point number or in @var{numerator}/@var{denominator} form.
4007 The sample rate of the incoming audio buffers.
4010 The sample format of the incoming audio buffers.
4011 Either a sample format name or its corresponding integer representation from
4012 the enum AVSampleFormat in @file{libavutil/samplefmt.h}
4014 @item channel_layout
4015 The channel layout of the incoming audio buffers.
4016 Either a channel layout name from channel_layout_map in
4017 @file{libavutil/channel_layout.c} or its corresponding integer representation
4018 from the AV_CH_LAYOUT_* macros in @file{libavutil/channel_layout.h}
4021 The number of channels of the incoming audio buffers.
4022 If both @var{channels} and @var{channel_layout} are specified, then they
4027 @subsection Examples
4030 abuffer=sample_rate=44100:sample_fmt=s16p:channel_layout=stereo
4033 will instruct the source to accept planar 16bit signed stereo at 44100Hz.
4034 Since the sample format with name "s16p" corresponds to the number
4035 6 and the "stereo" channel layout corresponds to the value 0x3, this is
4038 abuffer=sample_rate=44100:sample_fmt=6:channel_layout=0x3
4043 Generate an audio signal specified by an expression.
4045 This source accepts in input one or more expressions (one for each
4046 channel), which are evaluated and used to generate a corresponding
4049 This source accepts the following options:
4053 Set the '|'-separated expressions list for each separate channel. In case the
4054 @option{channel_layout} option is not specified, the selected channel layout
4055 depends on the number of provided expressions. Otherwise the last
4056 specified expression is applied to the remaining output channels.
4058 @item channel_layout, c
4059 Set the channel layout. The number of channels in the specified layout
4060 must be equal to the number of specified expressions.
4063 Set the minimum duration of the sourced audio. See
4064 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
4065 for the accepted syntax.
4066 Note that the resulting duration may be greater than the specified
4067 duration, as the generated audio is always cut at the end of a
4070 If not specified, or the expressed duration is negative, the audio is
4071 supposed to be generated forever.
4074 Set the number of samples per channel per each output frame,
4077 @item sample_rate, s
4078 Specify the sample rate, default to 44100.
4081 Each expression in @var{exprs} can contain the following constants:
4085 number of the evaluated sample, starting from 0
4088 time of the evaluated sample expressed in seconds, starting from 0
4095 @subsection Examples
4105 Generate a sin signal with frequency of 440 Hz, set sample rate to
4108 aevalsrc="sin(440*2*PI*t):s=8000"
4112 Generate a two channels signal, specify the channel layout (Front
4113 Center + Back Center) explicitly:
4115 aevalsrc="sin(420*2*PI*t)|cos(430*2*PI*t):c=FC|BC"
4119 Generate white noise:
4121 aevalsrc="-2+random(0)"
4125 Generate an amplitude modulated signal:
4127 aevalsrc="sin(10*2*PI*t)*sin(880*2*PI*t)"
4131 Generate 2.5 Hz binaural beats on a 360 Hz carrier:
4133 aevalsrc="0.1*sin(2*PI*(360-2.5/2)*t) | 0.1*sin(2*PI*(360+2.5/2)*t)"
4140 The null audio source, return unprocessed audio frames. It is mainly useful
4141 as a template and to be employed in analysis / debugging tools, or as
4142 the source for filters which ignore the input data (for example the sox
4145 This source accepts the following options:
4149 @item channel_layout, cl
4151 Specifies the channel layout, and can be either an integer or a string
4152 representing a channel layout. The default value of @var{channel_layout}
4155 Check the channel_layout_map definition in
4156 @file{libavutil/channel_layout.c} for the mapping between strings and
4157 channel layout values.
4159 @item sample_rate, r
4160 Specifies the sample rate, and defaults to 44100.
4163 Set the number of samples per requested frames.
4167 @subsection Examples
4171 Set the sample rate to 48000 Hz and the channel layout to AV_CH_LAYOUT_MONO.
4173 anullsrc=r=48000:cl=4
4177 Do the same operation with a more obvious syntax:
4179 anullsrc=r=48000:cl=mono
4183 All the parameters need to be explicitly defined.
4187 Synthesize a voice utterance using the libflite library.
4189 To enable compilation of this filter you need to configure FFmpeg with
4190 @code{--enable-libflite}.
4192 Note that the flite library is not thread-safe.
4194 The filter accepts the following options:
4199 If set to 1, list the names of the available voices and exit
4200 immediately. Default value is 0.
4203 Set the maximum number of samples per frame. Default value is 512.
4206 Set the filename containing the text to speak.
4209 Set the text to speak.
4212 Set the voice to use for the speech synthesis. Default value is
4213 @code{kal}. See also the @var{list_voices} option.
4216 @subsection Examples
4220 Read from file @file{speech.txt}, and synthesize the text using the
4221 standard flite voice:
4223 flite=textfile=speech.txt
4227 Read the specified text selecting the @code{slt} voice:
4229 flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4233 Input text to ffmpeg:
4235 ffmpeg -f lavfi -i flite=text='So fare thee well, poor devil of a Sub-Sub, whose commentator I am':voice=slt
4239 Make @file{ffplay} speak the specified text, using @code{flite} and
4240 the @code{lavfi} device:
4242 ffplay -f lavfi flite=text='No more be grieved for which that thou hast done.'
4246 For more information about libflite, check:
4247 @url{http://www.speech.cs.cmu.edu/flite/}
4251 Generate a noise audio signal.
4253 The filter accepts the following options:
4256 @item sample_rate, r
4257 Specify the sample rate. Default value is 48000 Hz.
4260 Specify the amplitude (0.0 - 1.0) of the generated audio stream. Default value
4264 Specify the duration of the generated audio stream. Not specifying this option
4265 results in noise with an infinite length.
4267 @item color, colour, c
4268 Specify the color of noise. Available noise colors are white, pink, and brown.
4269 Default color is white.
4272 Specify a value used to seed the PRNG.
4275 Set the number of samples per each output frame, default is 1024.
4278 @subsection Examples
4283 Generate 60 seconds of pink noise, with a 44.1 kHz sampling rate and an amplitude of 0.5:
4285 anoisesrc=d=60:c=pink:r=44100:a=0.5
4291 Generate an audio signal made of a sine wave with amplitude 1/8.
4293 The audio signal is bit-exact.
4295 The filter accepts the following options:
4300 Set the carrier frequency. Default is 440 Hz.
4302 @item beep_factor, b
4303 Enable a periodic beep every second with frequency @var{beep_factor} times
4304 the carrier frequency. Default is 0, meaning the beep is disabled.
4306 @item sample_rate, r
4307 Specify the sample rate, default is 44100.
4310 Specify the duration of the generated audio stream.
4312 @item samples_per_frame
4313 Set the number of samples per output frame.
4315 The expression can contain the following constants:
4319 The (sequential) number of the output audio frame, starting from 0.
4322 The PTS (Presentation TimeStamp) of the output audio frame,
4323 expressed in @var{TB} units.
4326 The PTS of the output audio frame, expressed in seconds.
4329 The timebase of the output audio frames.
4332 Default is @code{1024}.
4335 @subsection Examples
4340 Generate a simple 440 Hz sine wave:
4346 Generate a 220 Hz sine wave with a 880 Hz beep each second, for 5 seconds:
4350 sine=frequency=220:beep_factor=4:duration=5
4354 Generate a 1 kHz sine wave following @code{1602,1601,1602,1601,1602} NTSC
4357 sine=1000:samples_per_frame='st(0,mod(n,5)); 1602-not(not(eq(ld(0),1)+eq(ld(0),3)))'
4361 @c man end AUDIO SOURCES
4363 @chapter Audio Sinks
4364 @c man begin AUDIO SINKS
4366 Below is a description of the currently available audio sinks.
4368 @section abuffersink
4370 Buffer audio frames, and make them available to the end of filter chain.
4372 This sink is mainly intended for programmatic use, in particular
4373 through the interface defined in @file{libavfilter/buffersink.h}
4374 or the options system.
4376 It accepts a pointer to an AVABufferSinkContext structure, which
4377 defines the incoming buffers' formats, to be passed as the opaque
4378 parameter to @code{avfilter_init_filter} for initialization.
4381 Null audio sink; do absolutely nothing with the input audio. It is
4382 mainly useful as a template and for use in analysis / debugging
4385 @c man end AUDIO SINKS
4387 @chapter Video Filters
4388 @c man begin VIDEO FILTERS
4390 When you configure your FFmpeg build, you can disable any of the
4391 existing filters using @code{--disable-filters}.
4392 The configure output will show the video filters included in your
4395 Below is a description of the currently available video filters.
4397 @section alphaextract
4399 Extract the alpha component from the input as a grayscale video. This
4400 is especially useful with the @var{alphamerge} filter.
4404 Add or replace the alpha component of the primary input with the
4405 grayscale value of a second input. This is intended for use with
4406 @var{alphaextract} to allow the transmission or storage of frame
4407 sequences that have alpha in a format that doesn't support an alpha
4410 For example, to reconstruct full frames from a normal YUV-encoded video
4411 and a separate video created with @var{alphaextract}, you might use:
4413 movie=in_alpha.mkv [alpha]; [in][alpha] alphamerge [out]
4416 Since this filter is designed for reconstruction, it operates on frame
4417 sequences without considering timestamps, and terminates when either
4418 input reaches end of stream. This will cause problems if your encoding
4419 pipeline drops frames. If you're trying to apply an image as an
4420 overlay to a video stream, consider the @var{overlay} filter instead.
4424 Same as the @ref{subtitles} filter, except that it doesn't require libavcodec
4425 and libavformat to work. On the other hand, it is limited to ASS (Advanced
4426 Substation Alpha) subtitles files.
4428 This filter accepts the following option in addition to the common options from
4429 the @ref{subtitles} filter:
4433 Set the shaping engine
4435 Available values are:
4438 The default libass shaping engine, which is the best available.
4440 Fast, font-agnostic shaper that can do only substitutions
4442 Slower shaper using OpenType for substitutions and positioning
4445 The default is @code{auto}.
4449 Apply an Adaptive Temporal Averaging Denoiser to the video input.
4451 The filter accepts the following options:
4455 Set threshold A for 1st plane. Default is 0.02.
4456 Valid range is 0 to 0.3.
4459 Set threshold B for 1st plane. Default is 0.04.
4460 Valid range is 0 to 5.
4463 Set threshold A for 2nd plane. Default is 0.02.
4464 Valid range is 0 to 0.3.
4467 Set threshold B for 2nd plane. Default is 0.04.
4468 Valid range is 0 to 5.
4471 Set threshold A for 3rd plane. Default is 0.02.
4472 Valid range is 0 to 0.3.
4475 Set threshold B for 3rd plane. Default is 0.04.
4476 Valid range is 0 to 5.
4478 Threshold A is designed to react on abrupt changes in the input signal and
4479 threshold B is designed to react on continuous changes in the input signal.
4482 Set number of frames filter will use for averaging. Default is 33. Must be odd
4483 number in range [5, 129].
4486 Set what planes of frame filter will use for averaging. Default is all.
4491 Apply average blur filter.
4493 The filter accepts the following options:
4497 Set horizontal kernel size.
4500 Set which planes to filter. By default all planes are filtered.
4503 Set vertical kernel size, if zero it will be same as @code{sizeX}.
4504 Default is @code{0}.
4509 Compute the bounding box for the non-black pixels in the input frame
4512 This filter computes the bounding box containing all the pixels with a
4513 luminance value greater than the minimum allowed value.
4514 The parameters describing the bounding box are printed on the filter
4517 The filter accepts the following option:
4521 Set the minimal luminance value. Default is @code{16}.
4524 @section bitplanenoise
4526 Show and measure bit plane noise.
4528 The filter accepts the following options:
4532 Set which plane to analyze. Default is @code{1}.
4535 Filter out noisy pixels from @code{bitplane} set above.
4536 Default is disabled.
4539 @section blackdetect
4541 Detect video intervals that are (almost) completely black. Can be
4542 useful to detect chapter transitions, commercials, or invalid
4543 recordings. Output lines contains the time for the start, end and
4544 duration of the detected black interval expressed in seconds.
4546 In order to display the output lines, you need to set the loglevel at
4547 least to the AV_LOG_INFO value.
4549 The filter accepts the following options:
4552 @item black_min_duration, d
4553 Set the minimum detected black duration expressed in seconds. It must
4554 be a non-negative floating point number.
4556 Default value is 2.0.
4558 @item picture_black_ratio_th, pic_th
4559 Set the threshold for considering a picture "black".
4560 Express the minimum value for the ratio:
4562 @var{nb_black_pixels} / @var{nb_pixels}
4565 for which a picture is considered black.
4566 Default value is 0.98.
4568 @item pixel_black_th, pix_th
4569 Set the threshold for considering a pixel "black".
4571 The threshold expresses the maximum pixel luminance value for which a
4572 pixel is considered "black". The provided value is scaled according to
4573 the following equation:
4575 @var{absolute_threshold} = @var{luminance_minimum_value} + @var{pixel_black_th} * @var{luminance_range_size}
4578 @var{luminance_range_size} and @var{luminance_minimum_value} depend on
4579 the input video format, the range is [0-255] for YUV full-range
4580 formats and [16-235] for YUV non full-range formats.
4582 Default value is 0.10.
4585 The following example sets the maximum pixel threshold to the minimum
4586 value, and detects only black intervals of 2 or more seconds:
4588 blackdetect=d=2:pix_th=0.00
4593 Detect frames that are (almost) completely black. Can be useful to
4594 detect chapter transitions or commercials. Output lines consist of
4595 the frame number of the detected frame, the percentage of blackness,
4596 the position in the file if known or -1 and the timestamp in seconds.
4598 In order to display the output lines, you need to set the loglevel at
4599 least to the AV_LOG_INFO value.
4601 This filter exports frame metadata @code{lavfi.blackframe.pblack}.
4602 The value represents the percentage of pixels in the picture that
4603 are below the threshold value.
4605 It accepts the following parameters:
4610 The percentage of the pixels that have to be below the threshold; it defaults to
4613 @item threshold, thresh
4614 The threshold below which a pixel value is considered black; it defaults to
4619 @section blend, tblend
4621 Blend two video frames into each other.
4623 The @code{blend} filter takes two input streams and outputs one
4624 stream, the first input is the "top" layer and second input is
4625 "bottom" layer. By default, the output terminates when the longest input terminates.
4627 The @code{tblend} (time blend) filter takes two consecutive frames
4628 from one single stream, and outputs the result obtained by blending
4629 the new frame on top of the old frame.
4631 A description of the accepted options follows.
4639 Set blend mode for specific pixel component or all pixel components in case
4640 of @var{all_mode}. Default value is @code{normal}.
4642 Available values for component modes are:
4683 Set blend opacity for specific pixel component or all pixel components in case
4684 of @var{all_opacity}. Only used in combination with pixel component blend modes.
4691 Set blend expression for specific pixel component or all pixel components in case
4692 of @var{all_expr}. Note that related mode options will be ignored if those are set.
4694 The expressions can use the following variables:
4698 The sequential number of the filtered frame, starting from @code{0}.
4702 the coordinates of the current sample
4706 the width and height of currently filtered plane
4710 Width and height scale depending on the currently filtered plane. It is the
4711 ratio between the corresponding luma plane number of pixels and the current
4712 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
4713 @code{0.5,0.5} for chroma planes.
4716 Time of the current frame, expressed in seconds.
4719 Value of pixel component at current location for first video frame (top layer).
4722 Value of pixel component at current location for second video frame (bottom layer).
4726 Force termination when the shortest input terminates. Default is
4727 @code{0}. This option is only defined for the @code{blend} filter.
4730 Continue applying the last bottom frame after the end of the stream. A value of
4731 @code{0} disable the filter after the last frame of the bottom layer is reached.
4732 Default is @code{1}. This option is only defined for the @code{blend} filter.
4735 @subsection Examples
4739 Apply transition from bottom layer to top layer in first 10 seconds:
4741 blend=all_expr='A*(if(gte(T,10),1,T/10))+B*(1-(if(gte(T,10),1,T/10)))'
4745 Apply 1x1 checkerboard effect:
4747 blend=all_expr='if(eq(mod(X,2),mod(Y,2)),A,B)'
4751 Apply uncover left effect:
4753 blend=all_expr='if(gte(N*SW+X,W),A,B)'
4757 Apply uncover down effect:
4759 blend=all_expr='if(gte(Y-N*SH,0),A,B)'
4763 Apply uncover up-left effect:
4765 blend=all_expr='if(gte(T*SH*40+Y,H)*gte((T*40*SW+X)*W/H,W),A,B)'
4769 Split diagonally video and shows top and bottom layer on each side:
4771 blend=all_expr=if(gt(X,Y*(W/H)),A,B)
4775 Display differences between the current and the previous frame:
4777 tblend=all_mode=difference128
4783 Apply a boxblur algorithm to the input video.
4785 It accepts the following parameters:
4789 @item luma_radius, lr
4790 @item luma_power, lp
4791 @item chroma_radius, cr
4792 @item chroma_power, cp
4793 @item alpha_radius, ar
4794 @item alpha_power, ap
4798 A description of the accepted options follows.
4801 @item luma_radius, lr
4802 @item chroma_radius, cr
4803 @item alpha_radius, ar
4804 Set an expression for the box radius in pixels used for blurring the
4805 corresponding input plane.
4807 The radius value must be a non-negative number, and must not be
4808 greater than the value of the expression @code{min(w,h)/2} for the
4809 luma and alpha planes, and of @code{min(cw,ch)/2} for the chroma
4812 Default value for @option{luma_radius} is "2". If not specified,
4813 @option{chroma_radius} and @option{alpha_radius} default to the
4814 corresponding value set for @option{luma_radius}.
4816 The expressions can contain the following constants:
4820 The input width and height in pixels.
4824 The input chroma image width and height in pixels.
4828 The horizontal and vertical chroma subsample values. For example, for the
4829 pixel format "yuv422p", @var{hsub} is 2 and @var{vsub} is 1.
4832 @item luma_power, lp
4833 @item chroma_power, cp
4834 @item alpha_power, ap
4835 Specify how many times the boxblur filter is applied to the
4836 corresponding plane.
4838 Default value for @option{luma_power} is 2. If not specified,
4839 @option{chroma_power} and @option{alpha_power} default to the
4840 corresponding value set for @option{luma_power}.
4842 A value of 0 will disable the effect.
4845 @subsection Examples
4849 Apply a boxblur filter with the luma, chroma, and alpha radii
4852 boxblur=luma_radius=2:luma_power=1
4857 Set the luma radius to 2, and alpha and chroma radius to 0:
4859 boxblur=2:1:cr=0:ar=0
4863 Set the luma and chroma radii to a fraction of the video dimension:
4865 boxblur=luma_radius=min(h\,w)/10:luma_power=1:chroma_radius=min(cw\,ch)/10:chroma_power=1
4871 Deinterlace the input video ("bwdif" stands for "Bob Weaver
4872 Deinterlacing Filter").
4874 Motion adaptive deinterlacing based on yadif with the use of w3fdif and cubic
4875 interpolation algorithms.
4876 It accepts the following parameters:
4880 The interlacing mode to adopt. It accepts one of the following values:
4884 Output one frame for each frame.
4886 Output one frame for each field.
4889 The default value is @code{send_field}.
4892 The picture field parity assumed for the input interlaced video. It accepts one
4893 of the following values:
4897 Assume the top field is first.
4899 Assume the bottom field is first.
4901 Enable automatic detection of field parity.
4904 The default value is @code{auto}.
4905 If the interlacing is unknown or the decoder does not export this information,
4906 top field first will be assumed.
4909 Specify which frames to deinterlace. Accept one of the following
4914 Deinterlace all frames.
4916 Only deinterlace frames marked as interlaced.
4919 The default value is @code{all}.
4923 YUV colorspace color/chroma keying.
4925 The filter accepts the following options:
4929 The color which will be replaced with transparency.
4932 Similarity percentage with the key color.
4934 0.01 matches only the exact key color, while 1.0 matches everything.
4939 0.0 makes pixels either fully transparent, or not transparent at all.
4941 Higher values result in semi-transparent pixels, with a higher transparency
4942 the more similar the pixels color is to the key color.
4945 Signals that the color passed is already in YUV instead of RGB.
4947 Litteral colors like "green" or "red" don't make sense with this enabled anymore.
4948 This can be used to pass exact YUV values as hexadecimal numbers.
4951 @subsection Examples
4955 Make every green pixel in the input image transparent:
4957 ffmpeg -i input.png -vf chromakey=green out.png
4961 Overlay a greenscreen-video on top of a static black background.
4963 ffmpeg -f lavfi -i color=c=black:s=1280x720 -i video.mp4 -shortest -filter_complex "[1:v]chromakey=0x70de77:0.1:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.mkv
4969 Display CIE color diagram with pixels overlaid onto it.
4971 The filter accepts the following options:
4986 @item uhdtv, rec2020
4999 Set what gamuts to draw.
5001 See @code{system} option for available values.
5004 Set ciescope size, by default set to 512.
5007 Set intensity used to map input pixel values to CIE diagram.
5010 Set contrast used to draw tongue colors that are out of active color system gamut.
5013 Correct gamma displayed on scope, by default enabled.
5016 Show white point on CIE diagram, by default disabled.
5019 Set input gamma. Used only with XYZ input color space.
5024 Visualize information exported by some codecs.
5026 Some codecs can export information through frames using side-data or other
5027 means. For example, some MPEG based codecs export motion vectors through the
5028 @var{export_mvs} flag in the codec @option{flags2} option.
5030 The filter accepts the following option:
5034 Set motion vectors to visualize.
5036 Available flags for @var{mv} are:
5040 forward predicted MVs of P-frames
5042 forward predicted MVs of B-frames
5044 backward predicted MVs of B-frames
5048 Display quantization parameters using the chroma planes.
5051 Set motion vectors type to visualize. Includes MVs from all frames unless specified by @var{frame_type} option.
5053 Available flags for @var{mv_type} are:
5057 forward predicted MVs
5059 backward predicted MVs
5062 @item frame_type, ft
5063 Set frame type to visualize motion vectors of.
5065 Available flags for @var{frame_type} are:
5069 intra-coded frames (I-frames)
5071 predicted frames (P-frames)
5073 bi-directionally predicted frames (B-frames)
5077 @subsection Examples
5081 Visualize forward predicted MVs of all frames using @command{ffplay}:
5083 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv_type=fp
5087 Visualize multi-directionals MVs of P and B-Frames using @command{ffplay}:
5089 ffplay -flags2 +export_mvs input.mp4 -vf codecview=mv=pf+bf+bb
5093 @section colorbalance
5094 Modify intensity of primary colors (red, green and blue) of input frames.
5096 The filter allows an input frame to be adjusted in the shadows, midtones or highlights
5097 regions for the red-cyan, green-magenta or blue-yellow balance.
5099 A positive adjustment value shifts the balance towards the primary color, a negative
5100 value towards the complementary color.
5102 The filter accepts the following options:
5108 Adjust red, green and blue shadows (darkest pixels).
5113 Adjust red, green and blue midtones (medium pixels).
5118 Adjust red, green and blue highlights (brightest pixels).
5120 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5123 @subsection Examples
5127 Add red color cast to shadows:
5134 RGB colorspace color keying.
5136 The filter accepts the following options:
5140 The color which will be replaced with transparency.
5143 Similarity percentage with the key color.
5145 0.01 matches only the exact key color, while 1.0 matches everything.
5150 0.0 makes pixels either fully transparent, or not transparent at all.
5152 Higher values result in semi-transparent pixels, with a higher transparency
5153 the more similar the pixels color is to the key color.
5156 @subsection Examples
5160 Make every green pixel in the input image transparent:
5162 ffmpeg -i input.png -vf colorkey=green out.png
5166 Overlay a greenscreen-video on top of a static background image.
5168 ffmpeg -i background.png -i video.mp4 -filter_complex "[1:v]colorkey=0x3BBD1E:0.3:0.2[ckout];[0:v][ckout]overlay[out]" -map "[out]" output.flv
5172 @section colorlevels
5174 Adjust video input frames using levels.
5176 The filter accepts the following options:
5183 Adjust red, green, blue and alpha input black point.
5184 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{0}.
5190 Adjust red, green, blue and alpha input white point.
5191 Allowed ranges for options are @code{[-1.0, 1.0]}. Defaults are @code{1}.
5193 Input levels are used to lighten highlights (bright tones), darken shadows
5194 (dark tones), change the balance of bright and dark tones.
5200 Adjust red, green, blue and alpha output black point.
5201 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{0}.
5207 Adjust red, green, blue and alpha output white point.
5208 Allowed ranges for options are @code{[0, 1.0]}. Defaults are @code{1}.
5210 Output levels allows manual selection of a constrained output level range.
5213 @subsection Examples
5217 Make video output darker:
5219 colorlevels=rimin=0.058:gimin=0.058:bimin=0.058
5225 colorlevels=rimin=0.039:gimin=0.039:bimin=0.039:rimax=0.96:gimax=0.96:bimax=0.96
5229 Make video output lighter:
5231 colorlevels=rimax=0.902:gimax=0.902:bimax=0.902
5235 Increase brightness:
5237 colorlevels=romin=0.5:gomin=0.5:bomin=0.5
5241 @section colorchannelmixer
5243 Adjust video input frames by re-mixing color channels.
5245 This filter modifies a color channel by adding the values associated to
5246 the other channels of the same pixels. For example if the value to
5247 modify is red, the output value will be:
5249 @var{red}=@var{red}*@var{rr} + @var{blue}*@var{rb} + @var{green}*@var{rg} + @var{alpha}*@var{ra}
5252 The filter accepts the following options:
5259 Adjust contribution of input red, green, blue and alpha channels for output red channel.
5260 Default is @code{1} for @var{rr}, and @code{0} for @var{rg}, @var{rb} and @var{ra}.
5266 Adjust contribution of input red, green, blue and alpha channels for output green channel.
5267 Default is @code{1} for @var{gg}, and @code{0} for @var{gr}, @var{gb} and @var{ga}.
5273 Adjust contribution of input red, green, blue and alpha channels for output blue channel.
5274 Default is @code{1} for @var{bb}, and @code{0} for @var{br}, @var{bg} and @var{ba}.
5280 Adjust contribution of input red, green, blue and alpha channels for output alpha channel.
5281 Default is @code{1} for @var{aa}, and @code{0} for @var{ar}, @var{ag} and @var{ab}.
5283 Allowed ranges for options are @code{[-2.0, 2.0]}.
5286 @subsection Examples
5290 Convert source to grayscale:
5292 colorchannelmixer=.3:.4:.3:0:.3:.4:.3:0:.3:.4:.3
5295 Simulate sepia tones:
5297 colorchannelmixer=.393:.769:.189:0:.349:.686:.168:0:.272:.534:.131
5301 @section colormatrix
5303 Convert color matrix.
5305 The filter accepts the following options:
5310 Specify the source and destination color matrix. Both values must be
5313 The accepted values are:
5341 For example to convert from BT.601 to SMPTE-240M, use the command:
5343 colormatrix=bt601:smpte240m
5348 Convert colorspace, transfer characteristics or color primaries.
5349 Input video needs to have an even size.
5351 The filter accepts the following options:
5356 Specify all color properties at once.
5358 The accepted values are:
5388 Specify output colorspace.
5390 The accepted values are:
5399 BT.470BG or BT.601-6 625
5402 SMPTE-170M or BT.601-6 525
5411 BT.2020 with non-constant luminance
5417 Specify output transfer characteristics.
5419 The accepted values are:
5431 Constant gamma of 2.2
5434 Constant gamma of 2.8
5437 SMPTE-170M, BT.601-6 625 or BT.601-6 525
5455 BT.2020 for 10-bits content
5458 BT.2020 for 12-bits content
5464 Specify output color primaries.
5466 The accepted values are:
5475 BT.470BG or BT.601-6 625
5478 SMPTE-170M or BT.601-6 525
5499 Specify output color range.
5501 The accepted values are:
5504 TV (restricted) range
5507 MPEG (restricted) range
5518 Specify output color format.
5520 The accepted values are:
5523 YUV 4:2:0 planar 8-bits
5526 YUV 4:2:0 planar 10-bits
5529 YUV 4:2:0 planar 12-bits
5532 YUV 4:2:2 planar 8-bits
5535 YUV 4:2:2 planar 10-bits
5538 YUV 4:2:2 planar 12-bits
5541 YUV 4:4:4 planar 8-bits
5544 YUV 4:4:4 planar 10-bits
5547 YUV 4:4:4 planar 12-bits
5552 Do a fast conversion, which skips gamma/primary correction. This will take
5553 significantly less CPU, but will be mathematically incorrect. To get output
5554 compatible with that produced by the colormatrix filter, use fast=1.
5557 Specify dithering mode.
5559 The accepted values are:
5565 Floyd-Steinberg dithering
5569 Whitepoint adaptation mode.
5571 The accepted values are:
5574 Bradford whitepoint adaptation
5577 von Kries whitepoint adaptation
5580 identity whitepoint adaptation (i.e. no whitepoint adaptation)
5584 Override all input properties at once. Same accepted values as @ref{all}.
5587 Override input colorspace. Same accepted values as @ref{space}.
5590 Override input color primaries. Same accepted values as @ref{primaries}.
5593 Override input transfer characteristics. Same accepted values as @ref{trc}.
5596 Override input color range. Same accepted values as @ref{range}.
5600 The filter converts the transfer characteristics, color space and color
5601 primaries to the specified user values. The output value, if not specified,
5602 is set to a default value based on the "all" property. If that property is
5603 also not specified, the filter will log an error. The output color range and
5604 format default to the same value as the input color range and format. The
5605 input transfer characteristics, color space, color primaries and color range
5606 should be set on the input data. If any of these are missing, the filter will
5607 log an error and no conversion will take place.
5609 For example to convert the input to SMPTE-240M, use the command:
5611 colorspace=smpte240m
5614 @section convolution
5616 Apply convolution 3x3 or 5x5 filter.
5618 The filter accepts the following options:
5625 Set matrix for each plane.
5626 Matrix is sequence of 9 or 25 signed integers.
5632 Set multiplier for calculated value for each plane.
5638 Set bias for each plane. This value is added to the result of the multiplication.
5639 Useful for making the overall image brighter or darker. Default is 0.0.
5642 @subsection Examples
5648 convolution="0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0:0 -1 0 -1 5 -1 0 -1 0"
5654 convolution="1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1 1 1 1 1 1 1 1 1:1/9:1/9:1/9:1/9"
5660 convolution="0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:0 0 0 -1 1 0 0 0 0:5:1:1:1:0:128:128:128"
5666 convolution="0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:0 1 0 1 -4 1 0 1 0:5:5:5:1:0:128:128:128"
5672 convolution="-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2:-2 -1 0 -1 1 1 0 1 2"
5678 Copy the input video source unchanged to the output. This is mainly useful for
5683 Video filtering on GPU using Apple's CoreImage API on OSX.
5685 Hardware acceleration is based on an OpenGL context. Usually, this means it is
5686 processed by video hardware. However, software-based OpenGL implementations
5687 exist which means there is no guarantee for hardware processing. It depends on
5690 There are many filters and image generators provided by Apple that come with a
5691 large variety of options. The filter has to be referenced by its name along
5694 The coreimage filter accepts the following options:
5697 List all available filters and generators along with all their respective
5698 options as well as possible minimum and maximum values along with the default
5705 Specify all filters by their respective name and options.
5706 Use @var{list_filters} to determine all valid filter names and options.
5707 Numerical options are specified by a float value and are automatically clamped
5708 to their respective value range. Vector and color options have to be specified
5709 by a list of space separated float values. Character escaping has to be done.
5710 A special option name @code{default} is available to use default options for a
5713 It is required to specify either @code{default} or at least one of the filter options.
5714 All omitted options are used with their default values.
5715 The syntax of the filter string is as follows:
5717 filter=<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...][#<NAME>@@<OPTION>=<VALUE>[@@<OPTION>=<VALUE>][@@...]][#...]
5721 Specify a rectangle where the output of the filter chain is copied into the
5722 input image. It is given by a list of space separated float values:
5724 output_rect=x\ y\ width\ height
5726 If not given, the output rectangle equals the dimensions of the input image.
5727 The output rectangle is automatically cropped at the borders of the input
5728 image. Negative values are valid for each component.
5730 output_rect=25\ 25\ 100\ 100
5734 Several filters can be chained for successive processing without GPU-HOST
5735 transfers allowing for fast processing of complex filter chains.
5736 Currently, only filters with zero (generators) or exactly one (filters) input
5737 image and one output image are supported. Also, transition filters are not yet
5740 Some filters generate output images with additional padding depending on the
5741 respective filter kernel. The padding is automatically removed to ensure the
5742 filter output has the same size as the input image.
5744 For image generators, the size of the output image is determined by the
5745 previous output image of the filter chain or the input image of the whole
5746 filterchain, respectively. The generators do not use the pixel information of
5747 this image to generate their output. However, the generated output is
5748 blended onto this image, resulting in partial or complete coverage of the
5751 The @ref{coreimagesrc} video source can be used for generating input images
5752 which are directly fed into the filter chain. By using it, providing input
5753 images by another video source or an input video is not required.
5755 @subsection Examples
5760 List all filters available:
5762 coreimage=list_filters=true
5766 Use the CIBoxBlur filter with default options to blur an image:
5768 coreimage=filter=CIBoxBlur@@default
5772 Use a filter chain with CISepiaTone at default values and CIVignetteEffect with
5773 its center at 100x100 and a radius of 50 pixels:
5775 coreimage=filter=CIBoxBlur@@default#CIVignetteEffect@@inputCenter=100\ 100@@inputRadius=50
5779 Use nullsrc and CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
5780 given as complete and escaped command-line for Apple's standard bash shell:
5782 ffmpeg -f lavfi -i nullsrc=s=100x100,coreimage=filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
5788 Crop the input video to given dimensions.
5790 It accepts the following parameters:
5794 The width of the output video. It defaults to @code{iw}.
5795 This expression is evaluated only once during the filter
5796 configuration, or when the @samp{w} or @samp{out_w} command is sent.
5799 The height of the output video. It defaults to @code{ih}.
5800 This expression is evaluated only once during the filter
5801 configuration, or when the @samp{h} or @samp{out_h} command is sent.
5804 The horizontal position, in the input video, of the left edge of the output
5805 video. It defaults to @code{(in_w-out_w)/2}.
5806 This expression is evaluated per-frame.
5809 The vertical position, in the input video, of the top edge of the output video.
5810 It defaults to @code{(in_h-out_h)/2}.
5811 This expression is evaluated per-frame.
5814 If set to 1 will force the output display aspect ratio
5815 to be the same of the input, by changing the output sample aspect
5816 ratio. It defaults to 0.
5819 Enable exact cropping. If enabled, subsampled videos will be cropped at exact
5820 width/height/x/y as specified and will not be rounded to nearest smaller value.
5824 The @var{out_w}, @var{out_h}, @var{x}, @var{y} parameters are
5825 expressions containing the following constants:
5830 The computed values for @var{x} and @var{y}. They are evaluated for
5835 The input width and height.
5839 These are the same as @var{in_w} and @var{in_h}.
5843 The output (cropped) width and height.
5847 These are the same as @var{out_w} and @var{out_h}.
5850 same as @var{iw} / @var{ih}
5853 input sample aspect ratio
5856 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
5860 horizontal and vertical chroma subsample values. For example for the
5861 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
5864 The number of the input frame, starting from 0.
5867 the position in the file of the input frame, NAN if unknown
5870 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
5874 The expression for @var{out_w} may depend on the value of @var{out_h},
5875 and the expression for @var{out_h} may depend on @var{out_w}, but they
5876 cannot depend on @var{x} and @var{y}, as @var{x} and @var{y} are
5877 evaluated after @var{out_w} and @var{out_h}.
5879 The @var{x} and @var{y} parameters specify the expressions for the
5880 position of the top-left corner of the output (non-cropped) area. They
5881 are evaluated for each frame. If the evaluated value is not valid, it
5882 is approximated to the nearest valid value.
5884 The expression for @var{x} may depend on @var{y}, and the expression
5885 for @var{y} may depend on @var{x}.
5887 @subsection Examples
5891 Crop area with size 100x100 at position (12,34).
5896 Using named options, the example above becomes:
5898 crop=w=100:h=100:x=12:y=34
5902 Crop the central input area with size 100x100:
5908 Crop the central input area with size 2/3 of the input video:
5910 crop=2/3*in_w:2/3*in_h
5914 Crop the input video central square:
5921 Delimit the rectangle with the top-left corner placed at position
5922 100:100 and the right-bottom corner corresponding to the right-bottom
5923 corner of the input image.
5925 crop=in_w-100:in_h-100:100:100
5929 Crop 10 pixels from the left and right borders, and 20 pixels from
5930 the top and bottom borders
5932 crop=in_w-2*10:in_h-2*20
5936 Keep only the bottom right quarter of the input image:
5938 crop=in_w/2:in_h/2:in_w/2:in_h/2
5942 Crop height for getting Greek harmony:
5944 crop=in_w:1/PHI*in_w
5948 Apply trembling effect:
5950 crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(n/10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(n/7)
5954 Apply erratic camera effect depending on timestamp:
5956 crop=in_w/2:in_h/2:(in_w-out_w)/2+((in_w-out_w)/2)*sin(t*10):(in_h-out_h)/2 +((in_h-out_h)/2)*sin(t*13)"
5960 Set x depending on the value of y:
5962 crop=in_w/2:in_h/2:y:10+10*sin(n/10)
5966 @subsection Commands
5968 This filter supports the following commands:
5974 Set width/height of the output video and the horizontal/vertical position
5976 The command accepts the same syntax of the corresponding option.
5978 If the specified expression is not valid, it is kept at its current
5984 Auto-detect the crop size.
5986 It calculates the necessary cropping parameters and prints the
5987 recommended parameters via the logging system. The detected dimensions
5988 correspond to the non-black area of the input video.
5990 It accepts the following parameters:
5995 Set higher black value threshold, which can be optionally specified
5996 from nothing (0) to everything (255 for 8-bit based formats). An intensity
5997 value greater to the set value is considered non-black. It defaults to 24.
5998 You can also specify a value between 0.0 and 1.0 which will be scaled depending
5999 on the bitdepth of the pixel format.
6002 The value which the width/height should be divisible by. It defaults to
6003 16. The offset is automatically adjusted to center the video. Use 2 to
6004 get only even dimensions (needed for 4:2:2 video). 16 is best when
6005 encoding to most video codecs.
6007 @item reset_count, reset
6008 Set the counter that determines after how many frames cropdetect will
6009 reset the previously detected largest video area and start over to
6010 detect the current optimal crop area. Default value is 0.
6012 This can be useful when channel logos distort the video area. 0
6013 indicates 'never reset', and returns the largest area encountered during
6020 Apply color adjustments using curves.
6022 This filter is similar to the Adobe Photoshop and GIMP curves tools. Each
6023 component (red, green and blue) has its values defined by @var{N} key points
6024 tied from each other using a smooth curve. The x-axis represents the pixel
6025 values from the input frame, and the y-axis the new pixel values to be set for
6028 By default, a component curve is defined by the two points @var{(0;0)} and
6029 @var{(1;1)}. This creates a straight line where each original pixel value is
6030 "adjusted" to its own value, which means no change to the image.
6032 The filter allows you to redefine these two points and add some more. A new
6033 curve (using a natural cubic spline interpolation) will be define to pass
6034 smoothly through all these new coordinates. The new defined points needs to be
6035 strictly increasing over the x-axis, and their @var{x} and @var{y} values must
6036 be in the @var{[0;1]} interval. If the computed curves happened to go outside
6037 the vector spaces, the values will be clipped accordingly.
6039 The filter accepts the following options:
6043 Select one of the available color presets. This option can be used in addition
6044 to the @option{r}, @option{g}, @option{b} parameters; in this case, the later
6045 options takes priority on the preset values.
6046 Available presets are:
6049 @item color_negative
6052 @item increase_contrast
6054 @item linear_contrast
6055 @item medium_contrast
6057 @item strong_contrast
6060 Default is @code{none}.
6062 Set the master key points. These points will define a second pass mapping. It
6063 is sometimes called a "luminance" or "value" mapping. It can be used with
6064 @option{r}, @option{g}, @option{b} or @option{all} since it acts like a
6065 post-processing LUT.
6067 Set the key points for the red component.
6069 Set the key points for the green component.
6071 Set the key points for the blue component.
6073 Set the key points for all components (not including master).
6074 Can be used in addition to the other key points component
6075 options. In this case, the unset component(s) will fallback on this
6076 @option{all} setting.
6078 Specify a Photoshop curves file (@code{.acv}) to import the settings from.
6080 Save Gnuplot script of the curves in specified file.
6083 To avoid some filtergraph syntax conflicts, each key points list need to be
6084 defined using the following syntax: @code{x0/y0 x1/y1 x2/y2 ...}.
6086 @subsection Examples
6090 Increase slightly the middle level of blue:
6092 curves=blue='0/0 0.5/0.58 1/1'
6098 curves=r='0/0.11 .42/.51 1/0.95':g='0/0 0.50/0.48 1/1':b='0/0.22 .49/.44 1/0.8'
6100 Here we obtain the following coordinates for each components:
6103 @code{(0;0.11) (0.42;0.51) (1;0.95)}
6105 @code{(0;0) (0.50;0.48) (1;1)}
6107 @code{(0;0.22) (0.49;0.44) (1;0.80)}
6111 The previous example can also be achieved with the associated built-in preset:
6113 curves=preset=vintage
6123 Use a Photoshop preset and redefine the points of the green component:
6125 curves=psfile='MyCurvesPresets/purple.acv':green='0/0 0.45/0.53 1/1'
6129 Check out the curves of the @code{cross_process} profile using @command{ffmpeg}
6130 and @command{gnuplot}:
6132 ffmpeg -f lavfi -i color -vf curves=cross_process:plot=/tmp/curves.plt -frames:v 1 -f null -
6133 gnuplot -p /tmp/curves.plt
6139 Video data analysis filter.
6141 This filter shows hexadecimal pixel values of part of video.
6143 The filter accepts the following options:
6147 Set output video size.
6150 Set x offset from where to pick pixels.
6153 Set y offset from where to pick pixels.
6156 Set scope mode, can be one of the following:
6159 Draw hexadecimal pixel values with white color on black background.
6162 Draw hexadecimal pixel values with input video pixel color on black
6166 Draw hexadecimal pixel values on color background picked from input video,
6167 the text color is picked in such way so its always visible.
6171 Draw rows and columns numbers on left and top of video.
6174 Set background opacity.
6179 Denoise frames using 2D DCT (frequency domain filtering).
6181 This filter is not designed for real time.
6183 The filter accepts the following options:
6187 Set the noise sigma constant.
6189 This @var{sigma} defines a hard threshold of @code{3 * sigma}; every DCT
6190 coefficient (absolute value) below this threshold with be dropped.
6192 If you need a more advanced filtering, see @option{expr}.
6194 Default is @code{0}.
6197 Set number overlapping pixels for each block. Since the filter can be slow, you
6198 may want to reduce this value, at the cost of a less effective filter and the
6199 risk of various artefacts.
6201 If the overlapping value doesn't permit processing the whole input width or
6202 height, a warning will be displayed and according borders won't be denoised.
6204 Default value is @var{blocksize}-1, which is the best possible setting.
6207 Set the coefficient factor expression.
6209 For each coefficient of a DCT block, this expression will be evaluated as a
6210 multiplier value for the coefficient.
6212 If this is option is set, the @option{sigma} option will be ignored.
6214 The absolute value of the coefficient can be accessed through the @var{c}
6218 Set the @var{blocksize} using the number of bits. @code{1<<@var{n}} defines the
6219 @var{blocksize}, which is the width and height of the processed blocks.
6221 The default value is @var{3} (8x8) and can be raised to @var{4} for a
6222 @var{blocksize} of 16x16. Note that changing this setting has huge consequences
6223 on the speed processing. Also, a larger block size does not necessarily means a
6227 @subsection Examples
6229 Apply a denoise with a @option{sigma} of @code{4.5}:
6234 The same operation can be achieved using the expression system:
6236 dctdnoiz=e='gte(c, 4.5*3)'
6239 Violent denoise using a block size of @code{16x16}:
6246 Remove banding artifacts from input video.
6247 It works by replacing banded pixels with average value of referenced pixels.
6249 The filter accepts the following options:
6256 Set banding detection threshold for each plane. Default is 0.02.
6257 Valid range is 0.00003 to 0.5.
6258 If difference between current pixel and reference pixel is less than threshold,
6259 it will be considered as banded.
6262 Banding detection range in pixels. Default is 16. If positive, random number
6263 in range 0 to set value will be used. If negative, exact absolute value
6265 The range defines square of four pixels around current pixel.
6268 Set direction in radians from which four pixel will be compared. If positive,
6269 random direction from 0 to set direction will be picked. If negative, exact of
6270 absolute value will be picked. For example direction 0, -PI or -2*PI radians
6271 will pick only pixels on same row and -PI/2 will pick only pixels on same
6275 If enabled, current pixel is compared with average value of all four
6276 surrounding pixels. The default is enabled. If disabled current pixel is
6277 compared with all four surrounding pixels. The pixel is considered banded
6278 if only all four differences with surrounding pixels are less than threshold.
6281 If enabled, current pixel is changed if and only if all pixel components are banded,
6282 e.g. banding detection threshold is triggered for all color components.
6283 The default is disabled.
6289 Drop duplicated frames at regular intervals.
6291 The filter accepts the following options:
6295 Set the number of frames from which one will be dropped. Setting this to
6296 @var{N} means one frame in every batch of @var{N} frames will be dropped.
6297 Default is @code{5}.
6300 Set the threshold for duplicate detection. If the difference metric for a frame
6301 is less than or equal to this value, then it is declared as duplicate. Default
6305 Set scene change threshold. Default is @code{15}.
6309 Set the size of the x and y-axis blocks used during metric calculations.
6310 Larger blocks give better noise suppression, but also give worse detection of
6311 small movements. Must be a power of two. Default is @code{32}.
6314 Mark main input as a pre-processed input and activate clean source input
6315 stream. This allows the input to be pre-processed with various filters to help
6316 the metrics calculation while keeping the frame selection lossless. When set to
6317 @code{1}, the first stream is for the pre-processed input, and the second
6318 stream is the clean source from where the kept frames are chosen. Default is
6322 Set whether or not chroma is considered in the metric calculations. Default is
6328 Apply deflate effect to the video.
6330 This filter replaces the pixel by the local(3x3) average by taking into account
6331 only values lower than the pixel.
6333 It accepts the following options:
6340 Limit the maximum change for each plane, default is 65535.
6341 If 0, plane will remain unchanged.
6346 Remove temporal frame luminance variations.
6348 It accepts the following options:
6352 Set moving-average filter size in frames. Default is 5. Allowed range is 2 - 129.
6355 Set averaging mode to smooth temporal luminance variations.
6357 Available values are:
6384 Remove judder produced by partially interlaced telecined content.
6386 Judder can be introduced, for instance, by @ref{pullup} filter. If the original
6387 source was partially telecined content then the output of @code{pullup,dejudder}
6388 will have a variable frame rate. May change the recorded frame rate of the
6389 container. Aside from that change, this filter will not affect constant frame
6392 The option available in this filter is:
6396 Specify the length of the window over which the judder repeats.
6398 Accepts any integer greater than 1. Useful values are:
6402 If the original was telecined from 24 to 30 fps (Film to NTSC).
6405 If the original was telecined from 25 to 30 fps (PAL to NTSC).
6408 If a mixture of the two.
6411 The default is @samp{4}.
6416 Suppress a TV station logo by a simple interpolation of the surrounding
6417 pixels. Just set a rectangle covering the logo and watch it disappear
6418 (and sometimes something even uglier appear - your mileage may vary).
6420 It accepts the following parameters:
6425 Specify the top left corner coordinates of the logo. They must be
6430 Specify the width and height of the logo to clear. They must be
6434 Specify the thickness of the fuzzy edge of the rectangle (added to
6435 @var{w} and @var{h}). The default value is 1. This option is
6436 deprecated, setting higher values should no longer be necessary and
6440 When set to 1, a green rectangle is drawn on the screen to simplify
6441 finding the right @var{x}, @var{y}, @var{w}, and @var{h} parameters.
6442 The default value is 0.
6444 The rectangle is drawn on the outermost pixels which will be (partly)
6445 replaced with interpolated values. The values of the next pixels
6446 immediately outside this rectangle in each direction will be used to
6447 compute the interpolated pixel values inside the rectangle.
6451 @subsection Examples
6455 Set a rectangle covering the area with top left corner coordinates 0,0
6456 and size 100x77, and a band of size 10:
6458 delogo=x=0:y=0:w=100:h=77:band=10
6465 Attempt to fix small changes in horizontal and/or vertical shift. This
6466 filter helps remove camera shake from hand-holding a camera, bumping a
6467 tripod, moving on a vehicle, etc.
6469 The filter accepts the following options:
6477 Specify a rectangular area where to limit the search for motion
6479 If desired the search for motion vectors can be limited to a
6480 rectangular area of the frame defined by its top left corner, width
6481 and height. These parameters have the same meaning as the drawbox
6482 filter which can be used to visualise the position of the bounding
6485 This is useful when simultaneous movement of subjects within the frame
6486 might be confused for camera motion by the motion vector search.
6488 If any or all of @var{x}, @var{y}, @var{w} and @var{h} are set to -1
6489 then the full frame is used. This allows later options to be set
6490 without specifying the bounding box for the motion vector search.
6492 Default - search the whole frame.
6496 Specify the maximum extent of movement in x and y directions in the
6497 range 0-64 pixels. Default 16.
6500 Specify how to generate pixels to fill blanks at the edge of the
6501 frame. Available values are:
6504 Fill zeroes at blank locations
6506 Original image at blank locations
6508 Extruded edge value at blank locations
6510 Mirrored edge at blank locations
6512 Default value is @samp{mirror}.
6515 Specify the blocksize to use for motion search. Range 4-128 pixels,
6519 Specify the contrast threshold for blocks. Only blocks with more than
6520 the specified contrast (difference between darkest and lightest
6521 pixels) will be considered. Range 1-255, default 125.
6524 Specify the search strategy. Available values are:
6527 Set exhaustive search
6529 Set less exhaustive search.
6531 Default value is @samp{exhaustive}.
6534 If set then a detailed log of the motion search is written to the
6538 If set to 1, specify using OpenCL capabilities, only available if
6539 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
6545 Apply an exact inverse of the telecine operation. It requires a predefined
6546 pattern specified using the pattern option which must be the same as that passed
6547 to the telecine filter.
6549 This filter accepts the following options:
6558 The default value is @code{top}.
6562 A string of numbers representing the pulldown pattern you wish to apply.
6563 The default value is @code{23}.
6566 A number representing position of the first frame with respect to the telecine
6567 pattern. This is to be used if the stream is cut. The default value is @code{0}.
6572 Apply dilation effect to the video.
6574 This filter replaces the pixel by the local(3x3) maximum.
6576 It accepts the following options:
6583 Limit the maximum change for each plane, default is 65535.
6584 If 0, plane will remain unchanged.
6587 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
6590 Flags to local 3x3 coordinates maps like this:
6599 Displace pixels as indicated by second and third input stream.
6601 It takes three input streams and outputs one stream, the first input is the
6602 source, and second and third input are displacement maps.
6604 The second input specifies how much to displace pixels along the
6605 x-axis, while the third input specifies how much to displace pixels
6607 If one of displacement map streams terminates, last frame from that
6608 displacement map will be used.
6610 Note that once generated, displacements maps can be reused over and over again.
6612 A description of the accepted options follows.
6616 Set displace behavior for pixels that are out of range.
6618 Available values are:
6621 Missing pixels are replaced by black pixels.
6624 Adjacent pixels will spread out to replace missing pixels.
6627 Out of range pixels are wrapped so they point to pixels of other side.
6629 Default is @samp{smear}.
6633 @subsection Examples
6637 Add ripple effect to rgb input of video size hd720:
6639 ffmpeg -i INPUT -f lavfi -i nullsrc=s=hd720,lutrgb=128:128:128 -f lavfi -i nullsrc=s=hd720,geq='r=128+30*sin(2*PI*X/400+T):g=128+30*sin(2*PI*X/400+T):b=128+30*sin(2*PI*X/400+T)' -lavfi '[0][1][2]displace' OUTPUT
6643 Add wave effect to rgb input of video size hd720:
6645 ffmpeg -i INPUT -f lavfi -i nullsrc=hd720,geq='r=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):g=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T)):b=128+80*(sin(sqrt((X-W/2)*(X-W/2)+(Y-H/2)*(Y-H/2))/220*2*PI+T))' -lavfi '[1]split[x][y],[0][x][y]displace' OUTPUT
6651 Draw a colored box on the input image.
6653 It accepts the following parameters:
6658 The expressions which specify the top left corner coordinates of the box. It defaults to 0.
6662 The expressions which specify the width and height of the box; if 0 they are interpreted as
6663 the input width and height. It defaults to 0.
6666 Specify the color of the box to write. For the general syntax of this option,
6667 check the "Color" section in the ffmpeg-utils manual. If the special
6668 value @code{invert} is used, the box edge color is the same as the
6669 video with inverted luma.
6672 The expression which sets the thickness of the box edge. Default value is @code{3}.
6674 See below for the list of accepted constants.
6677 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6678 following constants:
6682 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6686 horizontal and vertical chroma subsample values. For example for the
6687 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6691 The input width and height.
6694 The input sample aspect ratio.
6698 The x and y offset coordinates where the box is drawn.
6702 The width and height of the drawn box.
6705 The thickness of the drawn box.
6707 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6708 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6712 @subsection Examples
6716 Draw a black box around the edge of the input image:
6722 Draw a box with color red and an opacity of 50%:
6724 drawbox=10:20:200:60:red@@0.5
6727 The previous example can be specified as:
6729 drawbox=x=10:y=20:w=200:h=60:color=red@@0.5
6733 Fill the box with pink color:
6735 drawbox=x=10:y=10:w=100:h=100:color=pink@@0.5:t=max
6739 Draw a 2-pixel red 2.40:1 mask:
6741 drawbox=x=-t:y=0.5*(ih-iw/2.4)-t:w=iw+t*2:h=iw/2.4+t*2:t=2:c=red
6747 Draw a grid on the input image.
6749 It accepts the following parameters:
6754 The expressions which specify the coordinates of some point of grid intersection (meant to configure offset). Both default to 0.
6758 The expressions which specify the width and height of the grid cell, if 0 they are interpreted as the
6759 input width and height, respectively, minus @code{thickness}, so image gets
6760 framed. Default to 0.
6763 Specify the color of the grid. For the general syntax of this option,
6764 check the "Color" section in the ffmpeg-utils manual. If the special
6765 value @code{invert} is used, the grid color is the same as the
6766 video with inverted luma.
6769 The expression which sets the thickness of the grid line. Default value is @code{1}.
6771 See below for the list of accepted constants.
6774 The parameters for @var{x}, @var{y}, @var{w} and @var{h} and @var{t} are expressions containing the
6775 following constants:
6779 The input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}.
6783 horizontal and vertical chroma subsample values. For example for the
6784 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
6788 The input grid cell width and height.
6791 The input sample aspect ratio.
6795 The x and y coordinates of some point of grid intersection (meant to configure offset).
6799 The width and height of the drawn cell.
6802 The thickness of the drawn cell.
6804 These constants allow the @var{x}, @var{y}, @var{w}, @var{h} and @var{t} expressions to refer to
6805 each other, so you may for example specify @code{y=x/dar} or @code{h=w/dar}.
6809 @subsection Examples
6813 Draw a grid with cell 100x100 pixels, thickness 2 pixels, with color red and an opacity of 50%:
6815 drawgrid=width=100:height=100:thickness=2:color=red@@0.5
6819 Draw a white 3x3 grid with an opacity of 50%:
6821 drawgrid=w=iw/3:h=ih/3:t=2:c=white@@0.5
6828 Draw a text string or text from a specified file on top of a video, using the
6829 libfreetype library.
6831 To enable compilation of this filter, you need to configure FFmpeg with
6832 @code{--enable-libfreetype}.
6833 To enable default font fallback and the @var{font} option you need to
6834 configure FFmpeg with @code{--enable-libfontconfig}.
6835 To enable the @var{text_shaping} option, you need to configure FFmpeg with
6836 @code{--enable-libfribidi}.
6840 It accepts the following parameters:
6845 Used to draw a box around text using the background color.
6846 The value must be either 1 (enable) or 0 (disable).
6847 The default value of @var{box} is 0.
6850 Set the width of the border to be drawn around the box using @var{boxcolor}.
6851 The default value of @var{boxborderw} is 0.
6854 The color to be used for drawing box around text. For the syntax of this
6855 option, check the "Color" section in the ffmpeg-utils manual.
6857 The default value of @var{boxcolor} is "white".
6860 Set the line spacing in pixels of the border to be drawn around the box using @var{box}.
6861 The default value of @var{line_spacing} is 0.
6864 Set the width of the border to be drawn around the text using @var{bordercolor}.
6865 The default value of @var{borderw} is 0.
6868 Set the color to be used for drawing border around text. For the syntax of this
6869 option, check the "Color" section in the ffmpeg-utils manual.
6871 The default value of @var{bordercolor} is "black".
6874 Select how the @var{text} is expanded. Can be either @code{none},
6875 @code{strftime} (deprecated) or
6876 @code{normal} (default). See the @ref{drawtext_expansion, Text expansion} section
6880 Set a start time for the count. Value is in microseconds. Only applied
6881 in the deprecated strftime expansion mode. To emulate in normal expansion
6882 mode use the @code{pts} function, supplying the start time (in seconds)
6883 as the second argument.
6886 If true, check and fix text coords to avoid clipping.
6889 The color to be used for drawing fonts. For the syntax of this option, check
6890 the "Color" section in the ffmpeg-utils manual.
6892 The default value of @var{fontcolor} is "black".
6894 @item fontcolor_expr
6895 String which is expanded the same way as @var{text} to obtain dynamic
6896 @var{fontcolor} value. By default this option has empty value and is not
6897 processed. When this option is set, it overrides @var{fontcolor} option.
6900 The font family to be used for drawing text. By default Sans.
6903 The font file to be used for drawing text. The path must be included.
6904 This parameter is mandatory if the fontconfig support is disabled.
6907 Draw the text applying alpha blending. The value can
6908 be a number between 0.0 and 1.0.
6909 The expression accepts the same variables @var{x, y} as well.
6910 The default value is 1.
6911 Please see @var{fontcolor_expr}.
6914 The font size to be used for drawing text.
6915 The default value of @var{fontsize} is 16.
6918 If set to 1, attempt to shape the text (for example, reverse the order of
6919 right-to-left text and join Arabic characters) before drawing it.
6920 Otherwise, just draw the text exactly as given.
6921 By default 1 (if supported).
6924 The flags to be used for loading the fonts.
6926 The flags map the corresponding flags supported by libfreetype, and are
6927 a combination of the following values:
6934 @item vertical_layout
6935 @item force_autohint
6938 @item ignore_global_advance_width
6940 @item ignore_transform
6946 Default value is "default".
6948 For more information consult the documentation for the FT_LOAD_*
6952 The color to be used for drawing a shadow behind the drawn text. For the
6953 syntax of this option, check the "Color" section in the ffmpeg-utils manual.
6955 The default value of @var{shadowcolor} is "black".
6959 The x and y offsets for the text shadow position with respect to the
6960 position of the text. They can be either positive or negative
6961 values. The default value for both is "0".
6964 The starting frame number for the n/frame_num variable. The default value
6968 The size in number of spaces to use for rendering the tab.
6972 Set the initial timecode representation in "hh:mm:ss[:;.]ff"
6973 format. It can be used with or without text parameter. @var{timecode_rate}
6974 option must be specified.
6976 @item timecode_rate, rate, r
6977 Set the timecode frame rate (timecode only).
6980 If set to 1, the output of the timecode option will wrap around at 24 hours.
6981 Default is 0 (disabled).
6984 The text string to be drawn. The text must be a sequence of UTF-8
6986 This parameter is mandatory if no file is specified with the parameter
6990 A text file containing text to be drawn. The text must be a sequence
6991 of UTF-8 encoded characters.
6993 This parameter is mandatory if no text string is specified with the
6994 parameter @var{text}.
6996 If both @var{text} and @var{textfile} are specified, an error is thrown.
6999 If set to 1, the @var{textfile} will be reloaded before each frame.
7000 Be sure to update it atomically, or it may be read partially, or even fail.
7004 The expressions which specify the offsets where text will be drawn
7005 within the video frame. They are relative to the top/left border of the
7008 The default value of @var{x} and @var{y} is "0".
7010 See below for the list of accepted constants and functions.
7013 The parameters for @var{x} and @var{y} are expressions containing the
7014 following constants and functions:
7018 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
7022 horizontal and vertical chroma subsample values. For example for the
7023 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
7026 the height of each text line
7034 @item max_glyph_a, ascent
7035 the maximum distance from the baseline to the highest/upper grid
7036 coordinate used to place a glyph outline point, for all the rendered
7038 It is a positive value, due to the grid's orientation with the Y axis
7041 @item max_glyph_d, descent
7042 the maximum distance from the baseline to the lowest grid coordinate
7043 used to place a glyph outline point, for all the rendered glyphs.
7044 This is a negative value, due to the grid's orientation, with the Y axis
7048 maximum glyph height, that is the maximum height for all the glyphs
7049 contained in the rendered text, it is equivalent to @var{ascent} -
7053 maximum glyph width, that is the maximum width for all the glyphs
7054 contained in the rendered text
7057 the number of input frame, starting from 0
7059 @item rand(min, max)
7060 return a random number included between @var{min} and @var{max}
7063 The input sample aspect ratio.
7066 timestamp expressed in seconds, NAN if the input timestamp is unknown
7069 the height of the rendered text
7072 the width of the rendered text
7076 the x and y offset coordinates where the text is drawn.
7078 These parameters allow the @var{x} and @var{y} expressions to refer
7079 each other, so you can for example specify @code{y=x/dar}.
7082 @anchor{drawtext_expansion}
7083 @subsection Text expansion
7085 If @option{expansion} is set to @code{strftime},
7086 the filter recognizes strftime() sequences in the provided text and
7087 expands them accordingly. Check the documentation of strftime(). This
7088 feature is deprecated.
7090 If @option{expansion} is set to @code{none}, the text is printed verbatim.
7092 If @option{expansion} is set to @code{normal} (which is the default),
7093 the following expansion mechanism is used.
7095 The backslash character @samp{\}, followed by any character, always expands to
7096 the second character.
7098 Sequences of the form @code{%@{...@}} are expanded. The text between the
7099 braces is a function name, possibly followed by arguments separated by ':'.
7100 If the arguments contain special characters or delimiters (':' or '@}'),
7101 they should be escaped.
7103 Note that they probably must also be escaped as the value for the
7104 @option{text} option in the filter argument string and as the filter
7105 argument in the filtergraph description, and possibly also for the shell,
7106 that makes up to four levels of escaping; using a text file avoids these
7109 The following functions are available:
7114 The expression evaluation result.
7116 It must take one argument specifying the expression to be evaluated,
7117 which accepts the same constants and functions as the @var{x} and
7118 @var{y} values. Note that not all constants should be used, for
7119 example the text size is not known when evaluating the expression, so
7120 the constants @var{text_w} and @var{text_h} will have an undefined
7123 @item expr_int_format, eif
7124 Evaluate the expression's value and output as formatted integer.
7126 The first argument is the expression to be evaluated, just as for the @var{expr} function.
7127 The second argument specifies the output format. Allowed values are @samp{x},
7128 @samp{X}, @samp{d} and @samp{u}. They are treated exactly as in the
7129 @code{printf} function.
7130 The third parameter is optional and sets the number of positions taken by the output.
7131 It can be used to add padding with zeros from the left.
7134 The time at which the filter is running, expressed in UTC.
7135 It can accept an argument: a strftime() format string.
7138 The time at which the filter is running, expressed in the local time zone.
7139 It can accept an argument: a strftime() format string.
7142 Frame metadata. Takes one or two arguments.
7144 The first argument is mandatory and specifies the metadata key.
7146 The second argument is optional and specifies a default value, used when the
7147 metadata key is not found or empty.
7150 The frame number, starting from 0.
7153 A 1 character description of the current picture type.
7156 The timestamp of the current frame.
7157 It can take up to three arguments.
7159 The first argument is the format of the timestamp; it defaults to @code{flt}
7160 for seconds as a decimal number with microsecond accuracy; @code{hms} stands
7161 for a formatted @var{[-]HH:MM:SS.mmm} timestamp with millisecond accuracy.
7162 @code{gmtime} stands for the timestamp of the frame formatted as UTC time;
7163 @code{localtime} stands for the timestamp of the frame formatted as
7164 local time zone time.
7166 The second argument is an offset added to the timestamp.
7168 If the format is set to @code{localtime} or @code{gmtime},
7169 a third argument may be supplied: a strftime() format string.
7170 By default, @var{YYYY-MM-DD HH:MM:SS} format will be used.
7173 @subsection Examples
7177 Draw "Test Text" with font FreeSerif, using the default values for the
7178 optional parameters.
7181 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text'"
7185 Draw 'Test Text' with font FreeSerif of size 24 at position x=100
7186 and y=50 (counting from the top-left corner of the screen), text is
7187 yellow with a red box around it. Both the text and the box have an
7191 drawtext="fontfile=/usr/share/fonts/truetype/freefont/FreeSerif.ttf: text='Test Text':\
7192 x=100: y=50: fontsize=24: fontcolor=yellow@@0.2: box=1: boxcolor=red@@0.2"
7195 Note that the double quotes are not necessary if spaces are not used
7196 within the parameter list.
7199 Show the text at the center of the video frame:
7201 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=(w-text_w)/2:y=(h-text_h)/2"
7205 Show the text at a random position, switching to a new position every 30 seconds:
7207 drawtext="fontsize=30:fontfile=FreeSerif.ttf:text='hello world':x=if(eq(mod(t\,30)\,0)\,rand(0\,(w-text_w))\,x):y=if(eq(mod(t\,30)\,0)\,rand(0\,(h-text_h))\,y)"
7211 Show a text line sliding from right to left in the last row of the video
7212 frame. The file @file{LONG_LINE} is assumed to contain a single line
7215 drawtext="fontsize=15:fontfile=FreeSerif.ttf:text=LONG_LINE:y=h-line_h:x=-50*t"
7219 Show the content of file @file{CREDITS} off the bottom of the frame and scroll up.
7221 drawtext="fontsize=20:fontfile=FreeSerif.ttf:textfile=CREDITS:y=h-20*t"
7225 Draw a single green letter "g", at the center of the input video.
7226 The glyph baseline is placed at half screen height.
7228 drawtext="fontsize=60:fontfile=FreeSerif.ttf:fontcolor=green:text=g:x=(w-max_glyph_w)/2:y=h/2-ascent"
7232 Show text for 1 second every 3 seconds:
7234 drawtext="fontfile=FreeSerif.ttf:fontcolor=white:x=100:y=x/dar:enable=lt(mod(t\,3)\,1):text='blink'"
7238 Use fontconfig to set the font. Note that the colons need to be escaped.
7240 drawtext='fontfile=Linux Libertine O-40\:style=Semibold:text=FFmpeg'
7244 Print the date of a real-time encoding (see strftime(3)):
7246 drawtext='fontfile=FreeSans.ttf:text=%@{localtime\:%a %b %d %Y@}'
7250 Show text fading in and out (appearing/disappearing):
7253 DS=1.0 # display start
7254 DE=10.0 # display end
7255 FID=1.5 # fade in duration
7256 FOD=5 # fade out duration
7257 ffplay -f lavfi "color,drawtext=text=TEST:fontsize=50:fontfile=FreeSerif.ttf:fontcolor_expr=ff0000%@{eif\\\\: clip(255*(1*between(t\\, $DS + $FID\\, $DE - $FOD) + ((t - $DS)/$FID)*between(t\\, $DS\\, $DS + $FID) + (-(t - $DE)/$FOD)*between(t\\, $DE - $FOD\\, $DE) )\\, 0\\, 255) \\\\: x\\\\: 2 @}"
7261 Horizontally align multiple separate texts. Note that @option{max_glyph_a}
7262 and the @option{fontsize} value are included in the @option{y} offset.
7264 drawtext=fontfile=FreeSans.ttf:text=DOG:fontsize=24:x=10:y=20+24-max_glyph_a,
7265 drawtext=fontfile=FreeSans.ttf:text=cow:fontsize=24:x=80:y=20+24-max_glyph_a
7270 For more information about libfreetype, check:
7271 @url{http://www.freetype.org/}.
7273 For more information about fontconfig, check:
7274 @url{http://freedesktop.org/software/fontconfig/fontconfig-user.html}.
7276 For more information about libfribidi, check:
7277 @url{http://fribidi.org/}.
7281 Detect and draw edges. The filter uses the Canny Edge Detection algorithm.
7283 The filter accepts the following options:
7288 Set low and high threshold values used by the Canny thresholding
7291 The high threshold selects the "strong" edge pixels, which are then
7292 connected through 8-connectivity with the "weak" edge pixels selected
7293 by the low threshold.
7295 @var{low} and @var{high} threshold values must be chosen in the range
7296 [0,1], and @var{low} should be lesser or equal to @var{high}.
7298 Default value for @var{low} is @code{20/255}, and default value for @var{high}
7302 Define the drawing mode.
7306 Draw white/gray wires on black background.
7309 Mix the colors to create a paint/cartoon effect.
7312 Default value is @var{wires}.
7315 @subsection Examples
7319 Standard edge detection with custom values for the hysteresis thresholding:
7321 edgedetect=low=0.1:high=0.4
7325 Painting effect without thresholding:
7327 edgedetect=mode=colormix:high=0
7332 Set brightness, contrast, saturation and approximate gamma adjustment.
7334 The filter accepts the following options:
7338 Set the contrast expression. The value must be a float value in range
7339 @code{-2.0} to @code{2.0}. The default value is "1".
7342 Set the brightness expression. The value must be a float value in
7343 range @code{-1.0} to @code{1.0}. The default value is "0".
7346 Set the saturation expression. The value must be a float in
7347 range @code{0.0} to @code{3.0}. The default value is "1".
7350 Set the gamma expression. The value must be a float in range
7351 @code{0.1} to @code{10.0}. The default value is "1".
7354 Set the gamma expression for red. The value must be a float in
7355 range @code{0.1} to @code{10.0}. The default value is "1".
7358 Set the gamma expression for green. The value must be a float in range
7359 @code{0.1} to @code{10.0}. The default value is "1".
7362 Set the gamma expression for blue. The value must be a float in range
7363 @code{0.1} to @code{10.0}. The default value is "1".
7366 Set the gamma weight expression. It can be used to reduce the effect
7367 of a high gamma value on bright image areas, e.g. keep them from
7368 getting overamplified and just plain white. The value must be a float
7369 in range @code{0.0} to @code{1.0}. A value of @code{0.0} turns the
7370 gamma correction all the way down while @code{1.0} leaves it at its
7371 full strength. Default is "1".
7374 Set when the expressions for brightness, contrast, saturation and
7375 gamma expressions are evaluated.
7377 It accepts the following values:
7380 only evaluate expressions once during the filter initialization or
7381 when a command is processed
7384 evaluate expressions for each incoming frame
7387 Default value is @samp{init}.
7390 The expressions accept the following parameters:
7393 frame count of the input frame starting from 0
7396 byte position of the corresponding packet in the input file, NAN if
7400 frame rate of the input video, NAN if the input frame rate is unknown
7403 timestamp expressed in seconds, NAN if the input timestamp is unknown
7406 @subsection Commands
7407 The filter supports the following commands:
7411 Set the contrast expression.
7414 Set the brightness expression.
7417 Set the saturation expression.
7420 Set the gamma expression.
7423 Set the gamma_r expression.
7426 Set gamma_g expression.
7429 Set gamma_b expression.
7432 Set gamma_weight expression.
7434 The command accepts the same syntax of the corresponding option.
7436 If the specified expression is not valid, it is kept at its current
7443 Apply erosion effect to the video.
7445 This filter replaces the pixel by the local(3x3) minimum.
7447 It accepts the following options:
7454 Limit the maximum change for each plane, default is 65535.
7455 If 0, plane will remain unchanged.
7458 Flag which specifies the pixel to refer to. Default is 255 i.e. all eight
7461 Flags to local 3x3 coordinates maps like this:
7468 @section extractplanes
7470 Extract color channel components from input video stream into
7471 separate grayscale video streams.
7473 The filter accepts the following option:
7477 Set plane(s) to extract.
7479 Available values for planes are:
7490 Choosing planes not available in the input will result in an error.
7491 That means you cannot select @code{r}, @code{g}, @code{b} planes
7492 with @code{y}, @code{u}, @code{v} planes at same time.
7495 @subsection Examples
7499 Extract luma, u and v color channel component from input video frame
7500 into 3 grayscale outputs:
7502 ffmpeg -i video.avi -filter_complex 'extractplanes=y+u+v[y][u][v]' -map '[y]' y.avi -map '[u]' u.avi -map '[v]' v.avi
7508 Apply a posterize effect using the ELBG (Enhanced LBG) algorithm.
7510 For each input image, the filter will compute the optimal mapping from
7511 the input to the output given the codebook length, that is the number
7512 of distinct output colors.
7514 This filter accepts the following options.
7517 @item codebook_length, l
7518 Set codebook length. The value must be a positive integer, and
7519 represents the number of distinct output colors. Default value is 256.
7522 Set the maximum number of iterations to apply for computing the optimal
7523 mapping. The higher the value the better the result and the higher the
7524 computation time. Default value is 1.
7527 Set a random seed, must be an integer included between 0 and
7528 UINT32_MAX. If not specified, or if explicitly set to -1, the filter
7529 will try to use a good random seed on a best effort basis.
7532 Set pal8 output pixel format. This option does not work with codebook
7533 length greater than 256.
7538 Apply a fade-in/out effect to the input video.
7540 It accepts the following parameters:
7544 The effect type can be either "in" for a fade-in, or "out" for a fade-out
7546 Default is @code{in}.
7548 @item start_frame, s
7549 Specify the number of the frame to start applying the fade
7550 effect at. Default is 0.
7553 The number of frames that the fade effect lasts. At the end of the
7554 fade-in effect, the output video will have the same intensity as the input video.
7555 At the end of the fade-out transition, the output video will be filled with the
7556 selected @option{color}.
7560 If set to 1, fade only alpha channel, if one exists on the input.
7563 @item start_time, st
7564 Specify the timestamp (in seconds) of the frame to start to apply the fade
7565 effect. If both start_frame and start_time are specified, the fade will start at
7566 whichever comes last. Default is 0.
7569 The number of seconds for which the fade effect has to last. At the end of the
7570 fade-in effect the output video will have the same intensity as the input video,
7571 at the end of the fade-out transition the output video will be filled with the
7572 selected @option{color}.
7573 If both duration and nb_frames are specified, duration is used. Default is 0
7574 (nb_frames is used by default).
7577 Specify the color of the fade. Default is "black".
7580 @subsection Examples
7584 Fade in the first 30 frames of video:
7589 The command above is equivalent to:
7595 Fade out the last 45 frames of a 200-frame video:
7598 fade=type=out:start_frame=155:nb_frames=45
7602 Fade in the first 25 frames and fade out the last 25 frames of a 1000-frame video:
7604 fade=in:0:25, fade=out:975:25
7608 Make the first 5 frames yellow, then fade in from frame 5-24:
7610 fade=in:5:20:color=yellow
7614 Fade in alpha over first 25 frames of video:
7616 fade=in:0:25:alpha=1
7620 Make the first 5.5 seconds black, then fade in for 0.5 seconds:
7622 fade=t=in:st=5.5:d=0.5
7628 Apply arbitrary expressions to samples in frequency domain
7632 Adjust the dc value (gain) of the luma plane of the image. The filter
7633 accepts an integer value in range @code{0} to @code{1000}. The default
7634 value is set to @code{0}.
7637 Adjust the dc value (gain) of the 1st chroma plane of the image. The
7638 filter accepts an integer value in range @code{0} to @code{1000}. The
7639 default value is set to @code{0}.
7642 Adjust the dc value (gain) of the 2nd chroma plane of the image. The
7643 filter accepts an integer value in range @code{0} to @code{1000}. The
7644 default value is set to @code{0}.
7647 Set the frequency domain weight expression for the luma plane.
7650 Set the frequency domain weight expression for the 1st chroma plane.
7653 Set the frequency domain weight expression for the 2nd chroma plane.
7655 The filter accepts the following variables:
7658 The coordinates of the current sample.
7662 The width and height of the image.
7665 @subsection Examples
7671 fftfilt=dc_Y=128:weight_Y='squish(1-(Y+X)/100)'
7677 fftfilt=dc_Y=0:weight_Y='squish((Y+X)/100-1)'
7683 fftfilt=dc_Y=0:weight_Y='1+squish(1-(Y+X)/100)'
7689 fftfilt=dc_Y=0:weight_Y='exp(-4 * ((Y+X)/(W+H)))'
7696 Extract a single field from an interlaced image using stride
7697 arithmetic to avoid wasting CPU time. The output frames are marked as
7700 The filter accepts the following options:
7704 Specify whether to extract the top (if the value is @code{0} or
7705 @code{top}) or the bottom field (if the value is @code{1} or
7711 Create new frames by copying the top and bottom fields from surrounding frames
7712 supplied as numbers by the hint file.
7716 Set file containing hints: absolute/relative frame numbers.
7718 There must be one line for each frame in a clip. Each line must contain two
7719 numbers separated by the comma, optionally followed by @code{-} or @code{+}.
7720 Numbers supplied on each line of file can not be out of [N-1,N+1] where N
7721 is current frame number for @code{absolute} mode or out of [-1, 1] range
7722 for @code{relative} mode. First number tells from which frame to pick up top
7723 field and second number tells from which frame to pick up bottom field.
7725 If optionally followed by @code{+} output frame will be marked as interlaced,
7726 else if followed by @code{-} output frame will be marked as progressive, else
7727 it will be marked same as input frame.
7728 If line starts with @code{#} or @code{;} that line is skipped.
7731 Can be item @code{absolute} or @code{relative}. Default is @code{absolute}.
7734 Example of first several lines of @code{hint} file for @code{relative} mode:
7737 1,0 - # second frame, use third's frame top field and second's frame bottom field
7738 1,0 - # third frame, use fourth's frame top field and third's frame bottom field
7755 Field matching filter for inverse telecine. It is meant to reconstruct the
7756 progressive frames from a telecined stream. The filter does not drop duplicated
7757 frames, so to achieve a complete inverse telecine @code{fieldmatch} needs to be
7758 followed by a decimation filter such as @ref{decimate} in the filtergraph.
7760 The separation of the field matching and the decimation is notably motivated by
7761 the possibility of inserting a de-interlacing filter fallback between the two.
7762 If the source has mixed telecined and real interlaced content,
7763 @code{fieldmatch} will not be able to match fields for the interlaced parts.
7764 But these remaining combed frames will be marked as interlaced, and thus can be
7765 de-interlaced by a later filter such as @ref{yadif} before decimation.
7767 In addition to the various configuration options, @code{fieldmatch} can take an
7768 optional second stream, activated through the @option{ppsrc} option. If
7769 enabled, the frames reconstruction will be based on the fields and frames from
7770 this second stream. This allows the first input to be pre-processed in order to
7771 help the various algorithms of the filter, while keeping the output lossless
7772 (assuming the fields are matched properly). Typically, a field-aware denoiser,
7773 or brightness/contrast adjustments can help.
7775 Note that this filter uses the same algorithms as TIVTC/TFM (AviSynth project)
7776 and VIVTC/VFM (VapourSynth project). The later is a light clone of TFM from
7777 which @code{fieldmatch} is based on. While the semantic and usage are very
7778 close, some behaviour and options names can differ.
7780 The @ref{decimate} filter currently only works for constant frame rate input.
7781 If your input has mixed telecined (30fps) and progressive content with a lower
7782 framerate like 24fps use the following filterchain to produce the necessary cfr
7783 stream: @code{dejudder,fps=30000/1001,fieldmatch,decimate}.
7785 The filter accepts the following options:
7789 Specify the assumed field order of the input stream. Available values are:
7793 Auto detect parity (use FFmpeg's internal parity value).
7795 Assume bottom field first.
7797 Assume top field first.
7800 Note that it is sometimes recommended not to trust the parity announced by the
7803 Default value is @var{auto}.
7806 Set the matching mode or strategy to use. @option{pc} mode is the safest in the
7807 sense that it won't risk creating jerkiness due to duplicate frames when
7808 possible, but if there are bad edits or blended fields it will end up
7809 outputting combed frames when a good match might actually exist. On the other
7810 hand, @option{pcn_ub} mode is the most risky in terms of creating jerkiness,
7811 but will almost always find a good frame if there is one. The other values are
7812 all somewhere in between @option{pc} and @option{pcn_ub} in terms of risking
7813 jerkiness and creating duplicate frames versus finding good matches in sections
7814 with bad edits, orphaned fields, blended fields, etc.
7816 More details about p/c/n/u/b are available in @ref{p/c/n/u/b meaning} section.
7818 Available values are:
7822 2-way matching (p/c)
7824 2-way matching, and trying 3rd match if still combed (p/c + n)
7826 2-way matching, and trying 3rd match (same order) if still combed (p/c + u)
7828 2-way matching, trying 3rd match if still combed, and trying 4th/5th matches if
7829 still combed (p/c + n + u/b)
7831 3-way matching (p/c/n)
7833 3-way matching, and trying 4th/5th matches if all 3 of the original matches are
7834 detected as combed (p/c/n + u/b)
7837 The parenthesis at the end indicate the matches that would be used for that
7838 mode assuming @option{order}=@var{tff} (and @option{field} on @var{auto} or
7841 In terms of speed @option{pc} mode is by far the fastest and @option{pcn_ub} is
7844 Default value is @var{pc_n}.
7847 Mark the main input stream as a pre-processed input, and enable the secondary
7848 input stream as the clean source to pick the fields from. See the filter
7849 introduction for more details. It is similar to the @option{clip2} feature from
7852 Default value is @code{0} (disabled).
7855 Set the field to match from. It is recommended to set this to the same value as
7856 @option{order} unless you experience matching failures with that setting. In
7857 certain circumstances changing the field that is used to match from can have a
7858 large impact on matching performance. Available values are:
7862 Automatic (same value as @option{order}).
7864 Match from the bottom field.
7866 Match from the top field.
7869 Default value is @var{auto}.
7872 Set whether or not chroma is included during the match comparisons. In most
7873 cases it is recommended to leave this enabled. You should set this to @code{0}
7874 only if your clip has bad chroma problems such as heavy rainbowing or other
7875 artifacts. Setting this to @code{0} could also be used to speed things up at
7876 the cost of some accuracy.
7878 Default value is @code{1}.
7882 These define an exclusion band which excludes the lines between @option{y0} and
7883 @option{y1} from being included in the field matching decision. An exclusion
7884 band can be used to ignore subtitles, a logo, or other things that may
7885 interfere with the matching. @option{y0} sets the starting scan line and
7886 @option{y1} sets the ending line; all lines in between @option{y0} and
7887 @option{y1} (including @option{y0} and @option{y1}) will be ignored. Setting
7888 @option{y0} and @option{y1} to the same value will disable the feature.
7889 @option{y0} and @option{y1} defaults to @code{0}.
7892 Set the scene change detection threshold as a percentage of maximum change on
7893 the luma plane. Good values are in the @code{[8.0, 14.0]} range. Scene change
7894 detection is only relevant in case @option{combmatch}=@var{sc}. The range for
7895 @option{scthresh} is @code{[0.0, 100.0]}.
7897 Default value is @code{12.0}.
7900 When @option{combatch} is not @var{none}, @code{fieldmatch} will take into
7901 account the combed scores of matches when deciding what match to use as the
7902 final match. Available values are:
7906 No final matching based on combed scores.
7908 Combed scores are only used when a scene change is detected.
7910 Use combed scores all the time.
7913 Default is @var{sc}.
7916 Force @code{fieldmatch} to calculate the combed metrics for certain matches and
7917 print them. This setting is known as @option{micout} in TFM/VFM vocabulary.
7918 Available values are:
7922 No forced calculation.
7924 Force p/c/n calculations.
7926 Force p/c/n/u/b calculations.
7929 Default value is @var{none}.
7932 This is the area combing threshold used for combed frame detection. This
7933 essentially controls how "strong" or "visible" combing must be to be detected.
7934 Larger values mean combing must be more visible and smaller values mean combing
7935 can be less visible or strong and still be detected. Valid settings are from
7936 @code{-1} (every pixel will be detected as combed) to @code{255} (no pixel will
7937 be detected as combed). This is basically a pixel difference value. A good
7938 range is @code{[8, 12]}.
7940 Default value is @code{9}.
7943 Sets whether or not chroma is considered in the combed frame decision. Only
7944 disable this if your source has chroma problems (rainbowing, etc.) that are
7945 causing problems for the combed frame detection with chroma enabled. Actually,
7946 using @option{chroma}=@var{0} is usually more reliable, except for the case
7947 where there is chroma only combing in the source.
7949 Default value is @code{0}.
7953 Respectively set the x-axis and y-axis size of the window used during combed
7954 frame detection. This has to do with the size of the area in which
7955 @option{combpel} pixels are required to be detected as combed for a frame to be
7956 declared combed. See the @option{combpel} parameter description for more info.
7957 Possible values are any number that is a power of 2 starting at 4 and going up
7960 Default value is @code{16}.
7963 The number of combed pixels inside any of the @option{blocky} by
7964 @option{blockx} size blocks on the frame for the frame to be detected as
7965 combed. While @option{cthresh} controls how "visible" the combing must be, this
7966 setting controls "how much" combing there must be in any localized area (a
7967 window defined by the @option{blockx} and @option{blocky} settings) on the
7968 frame. Minimum value is @code{0} and maximum is @code{blocky x blockx} (at
7969 which point no frames will ever be detected as combed). This setting is known
7970 as @option{MI} in TFM/VFM vocabulary.
7972 Default value is @code{80}.
7975 @anchor{p/c/n/u/b meaning}
7976 @subsection p/c/n/u/b meaning
7978 @subsubsection p/c/n
7980 We assume the following telecined stream:
7983 Top fields: 1 2 2 3 4
7984 Bottom fields: 1 2 3 4 4
7987 The numbers correspond to the progressive frame the fields relate to. Here, the
7988 first two frames are progressive, the 3rd and 4th are combed, and so on.
7990 When @code{fieldmatch} is configured to run a matching from bottom
7991 (@option{field}=@var{bottom}) this is how this input stream get transformed:
7996 B 1 2 3 4 4 <-- matching reference
8005 As a result of the field matching, we can see that some frames get duplicated.
8006 To perform a complete inverse telecine, you need to rely on a decimation filter
8007 after this operation. See for instance the @ref{decimate} filter.
8009 The same operation now matching from top fields (@option{field}=@var{top})
8014 T 1 2 2 3 4 <-- matching reference
8024 In these examples, we can see what @var{p}, @var{c} and @var{n} mean;
8025 basically, they refer to the frame and field of the opposite parity:
8028 @item @var{p} matches the field of the opposite parity in the previous frame
8029 @item @var{c} matches the field of the opposite parity in the current frame
8030 @item @var{n} matches the field of the opposite parity in the next frame
8035 The @var{u} and @var{b} matching are a bit special in the sense that they match
8036 from the opposite parity flag. In the following examples, we assume that we are
8037 currently matching the 2nd frame (Top:2, bottom:2). According to the match, a
8038 'x' is placed above and below each matched fields.
8040 With bottom matching (@option{field}=@var{bottom}):
8045 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8046 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8054 With top matching (@option{field}=@var{top}):
8059 Top 1 2 2 1 2 2 1 2 2 1 2 2 1 2 2
8060 Bottom 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3
8068 @subsection Examples
8070 Simple IVTC of a top field first telecined stream:
8072 fieldmatch=order=tff:combmatch=none, decimate
8075 Advanced IVTC, with fallback on @ref{yadif} for still combed frames:
8077 fieldmatch=order=tff:combmatch=full, yadif=deint=interlaced, decimate
8082 Transform the field order of the input video.
8084 It accepts the following parameters:
8089 The output field order. Valid values are @var{tff} for top field first or @var{bff}
8090 for bottom field first.
8093 The default value is @samp{tff}.
8095 The transformation is done by shifting the picture content up or down
8096 by one line, and filling the remaining line with appropriate picture content.
8097 This method is consistent with most broadcast field order converters.
8099 If the input video is not flagged as being interlaced, or it is already
8100 flagged as being of the required output field order, then this filter does
8101 not alter the incoming video.
8103 It is very useful when converting to or from PAL DV material,
8104 which is bottom field first.
8108 ffmpeg -i in.vob -vf "fieldorder=bff" out.dv
8111 @section fifo, afifo
8113 Buffer input images and send them when they are requested.
8115 It is mainly useful when auto-inserted by the libavfilter
8118 It does not take parameters.
8122 Find a rectangular object
8124 It accepts the following options:
8128 Filepath of the object image, needs to be in gray8.
8131 Detection threshold, default is 0.5.
8134 Number of mipmaps, default is 3.
8136 @item xmin, ymin, xmax, ymax
8137 Specifies the rectangle in which to search.
8140 @subsection Examples
8144 Generate a representative palette of a given video using @command{ffmpeg}:
8146 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8152 Cover a rectangular object
8154 It accepts the following options:
8158 Filepath of the optional cover image, needs to be in yuv420.
8163 It accepts the following values:
8166 cover it by the supplied image
8168 cover it by interpolating the surrounding pixels
8171 Default value is @var{blur}.
8174 @subsection Examples
8178 Generate a representative palette of a given video using @command{ffmpeg}:
8180 ffmpeg -i file.ts -vf find_rect=newref.pgm,cover_rect=cover.jpg:mode=cover new.mkv
8187 Convert the input video to one of the specified pixel formats.
8188 Libavfilter will try to pick one that is suitable as input to
8191 It accepts the following parameters:
8195 A '|'-separated list of pixel format names, such as
8196 "pix_fmts=yuv420p|monow|rgb24".
8200 @subsection Examples
8204 Convert the input video to the @var{yuv420p} format
8206 format=pix_fmts=yuv420p
8209 Convert the input video to any of the formats in the list
8211 format=pix_fmts=yuv420p|yuv444p|yuv410p
8218 Convert the video to specified constant frame rate by duplicating or dropping
8219 frames as necessary.
8221 It accepts the following parameters:
8225 The desired output frame rate. The default is @code{25}.
8230 Possible values are:
8233 zero round towards 0
8237 round towards -infinity
8239 round towards +infinity
8243 The default is @code{near}.
8246 Assume the first PTS should be the given value, in seconds. This allows for
8247 padding/trimming at the start of stream. By default, no assumption is made
8248 about the first frame's expected PTS, so no padding or trimming is done.
8249 For example, this could be set to 0 to pad the beginning with duplicates of
8250 the first frame if a video stream starts after the audio stream or to trim any
8251 frames with a negative PTS.
8255 Alternatively, the options can be specified as a flat string:
8256 @var{fps}[:@var{round}].
8258 See also the @ref{setpts} filter.
8260 @subsection Examples
8264 A typical usage in order to set the fps to 25:
8270 Sets the fps to 24, using abbreviation and rounding method to round to nearest:
8272 fps=fps=film:round=near
8278 Pack two different video streams into a stereoscopic video, setting proper
8279 metadata on supported codecs. The two views should have the same size and
8280 framerate and processing will stop when the shorter video ends. Please note
8281 that you may conveniently adjust view properties with the @ref{scale} and
8284 It accepts the following parameters:
8288 The desired packing format. Supported values are:
8293 The views are next to each other (default).
8296 The views are on top of each other.
8299 The views are packed by line.
8302 The views are packed by column.
8305 The views are temporally interleaved.
8314 # Convert left and right views into a frame-sequential video
8315 ffmpeg -i LEFT -i RIGHT -filter_complex framepack=frameseq OUTPUT
8317 # Convert views into a side-by-side video with the same output resolution as the input
8318 ffmpeg -i LEFT -i RIGHT -filter_complex [0:v]scale=w=iw/2[left],[1:v]scale=w=iw/2[right],[left][right]framepack=sbs OUTPUT
8323 Change the frame rate by interpolating new video output frames from the source
8326 This filter is not designed to function correctly with interlaced media. If
8327 you wish to change the frame rate of interlaced media then you are required
8328 to deinterlace before this filter and re-interlace after this filter.
8330 A description of the accepted options follows.
8334 Specify the output frames per second. This option can also be specified
8335 as a value alone. The default is @code{50}.
8338 Specify the start of a range where the output frame will be created as a
8339 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8340 the default is @code{15}.
8343 Specify the end of a range where the output frame will be created as a
8344 linear interpolation of two frames. The range is [@code{0}-@code{255}],
8345 the default is @code{240}.
8348 Specify the level at which a scene change is detected as a value between
8349 0 and 100 to indicate a new scene; a low value reflects a low
8350 probability for the current frame to introduce a new scene, while a higher
8351 value means the current frame is more likely to be one.
8352 The default is @code{7}.
8355 Specify flags influencing the filter process.
8357 Available value for @var{flags} is:
8360 @item scene_change_detect, scd
8361 Enable scene change detection using the value of the option @var{scene}.
8362 This flag is enabled by default.
8368 Select one frame every N-th frame.
8370 This filter accepts the following option:
8373 Select frame after every @code{step} frames.
8374 Allowed values are positive integers higher than 0. Default value is @code{1}.
8380 Apply a frei0r effect to the input video.
8382 To enable the compilation of this filter, you need to install the frei0r
8383 header and configure FFmpeg with @code{--enable-frei0r}.
8385 It accepts the following parameters:
8390 The name of the frei0r effect to load. If the environment variable
8391 @env{FREI0R_PATH} is defined, the frei0r effect is searched for in each of the
8392 directories specified by the colon-separated list in @env{FREIOR_PATH}.
8393 Otherwise, the standard frei0r paths are searched, in this order:
8394 @file{HOME/.frei0r-1/lib/}, @file{/usr/local/lib/frei0r-1/},
8395 @file{/usr/lib/frei0r-1/}.
8398 A '|'-separated list of parameters to pass to the frei0r effect.
8402 A frei0r effect parameter can be a boolean (its value is either
8403 "y" or "n"), a double, a color (specified as
8404 @var{R}/@var{G}/@var{B}, where @var{R}, @var{G}, and @var{B} are floating point
8405 numbers between 0.0 and 1.0, inclusive) or by a color description specified in the "Color"
8406 section in the ffmpeg-utils manual), a position (specified as @var{X}/@var{Y}, where
8407 @var{X} and @var{Y} are floating point numbers) and/or a string.
8409 The number and types of parameters depend on the loaded effect. If an
8410 effect parameter is not specified, the default value is set.
8412 @subsection Examples
8416 Apply the distort0r effect, setting the first two double parameters:
8418 frei0r=filter_name=distort0r:filter_params=0.5|0.01
8422 Apply the colordistance effect, taking a color as the first parameter:
8424 frei0r=colordistance:0.2/0.3/0.4
8425 frei0r=colordistance:violet
8426 frei0r=colordistance:0x112233
8430 Apply the perspective effect, specifying the top left and top right image
8433 frei0r=perspective:0.2/0.2|0.8/0.2
8437 For more information, see
8438 @url{http://frei0r.dyne.org}
8442 Apply fast and simple postprocessing. It is a faster version of @ref{spp}.
8444 It splits (I)DCT into horizontal/vertical passes. Unlike the simple post-
8445 processing filter, one of them is performed once per block, not per pixel.
8446 This allows for much higher speed.
8448 The filter accepts the following options:
8452 Set quality. This option defines the number of levels for averaging. It accepts
8453 an integer in the range 4-5. Default value is @code{4}.
8456 Force a constant quantization parameter. It accepts an integer in range 0-63.
8457 If not set, the filter will use the QP from the video stream (if available).
8460 Set filter strength. It accepts an integer in range -15 to 32. Lower values mean
8461 more details but also more artifacts, while higher values make the image smoother
8462 but also blurrier. Default value is @code{0} − PSNR optimal.
8465 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
8466 option may cause flicker since the B-Frames have often larger QP. Default is
8467 @code{0} (not enabled).
8473 Apply Gaussian blur filter.
8475 The filter accepts the following options:
8479 Set horizontal sigma, standard deviation of Gaussian blur. Default is @code{0.5}.
8482 Set number of steps for Gaussian approximation. Defauls is @code{1}.
8485 Set which planes to filter. By default all planes are filtered.
8488 Set vertical sigma, if negative it will be same as @code{sigma}.
8489 Default is @code{-1}.
8494 The filter accepts the following options:
8498 Set the luminance expression.
8500 Set the chrominance blue expression.
8502 Set the chrominance red expression.
8504 Set the alpha expression.
8506 Set the red expression.
8508 Set the green expression.
8510 Set the blue expression.
8513 The colorspace is selected according to the specified options. If one
8514 of the @option{lum_expr}, @option{cb_expr}, or @option{cr_expr}
8515 options is specified, the filter will automatically select a YCbCr
8516 colorspace. If one of the @option{red_expr}, @option{green_expr}, or
8517 @option{blue_expr} options is specified, it will select an RGB
8520 If one of the chrominance expression is not defined, it falls back on the other
8521 one. If no alpha expression is specified it will evaluate to opaque value.
8522 If none of chrominance expressions are specified, they will evaluate
8523 to the luminance expression.
8525 The expressions can use the following variables and functions:
8529 The sequential number of the filtered frame, starting from @code{0}.
8533 The coordinates of the current sample.
8537 The width and height of the image.
8541 Width and height scale depending on the currently filtered plane. It is the
8542 ratio between the corresponding luma plane number of pixels and the current
8543 plane ones. E.g. for YUV4:2:0 the values are @code{1,1} for the luma plane, and
8544 @code{0.5,0.5} for chroma planes.
8547 Time of the current frame, expressed in seconds.
8550 Return the value of the pixel at location (@var{x},@var{y}) of the current
8554 Return the value of the pixel at location (@var{x},@var{y}) of the luminance
8558 Return the value of the pixel at location (@var{x},@var{y}) of the
8559 blue-difference chroma plane. Return 0 if there is no such plane.
8562 Return the value of the pixel at location (@var{x},@var{y}) of the
8563 red-difference chroma plane. Return 0 if there is no such plane.
8568 Return the value of the pixel at location (@var{x},@var{y}) of the
8569 red/green/blue component. Return 0 if there is no such component.
8572 Return the value of the pixel at location (@var{x},@var{y}) of the alpha
8573 plane. Return 0 if there is no such plane.
8576 For functions, if @var{x} and @var{y} are outside the area, the value will be
8577 automatically clipped to the closer edge.
8579 @subsection Examples
8583 Flip the image horizontally:
8589 Generate a bidimensional sine wave, with angle @code{PI/3} and a
8590 wavelength of 100 pixels:
8592 geq=128 + 100*sin(2*(PI/100)*(cos(PI/3)*(X-50*T) + sin(PI/3)*Y)):128:128
8596 Generate a fancy enigmatic moving light:
8598 nullsrc=s=256x256,geq=random(1)/hypot(X-cos(N*0.07)*W/2-W/2\,Y-sin(N*0.09)*H/2-H/2)^2*1000000*sin(N*0.02):128:128
8602 Generate a quick emboss effect:
8604 format=gray,geq=lum_expr='(p(X,Y)+(256-p(X-4,Y-4)))/2'
8608 Modify RGB components depending on pixel position:
8610 geq=r='X/W*r(X,Y)':g='(1-X/W)*g(X,Y)':b='(H-Y)/H*b(X,Y)'
8614 Create a radial gradient that is the same size as the input (also see
8615 the @ref{vignette} filter):
8617 geq=lum=255*gauss((X/W-0.5)*3)*gauss((Y/H-0.5)*3)/gauss(0)/gauss(0),format=gray
8623 Fix the banding artifacts that are sometimes introduced into nearly flat
8624 regions by truncation to 8-bit color depth.
8625 Interpolate the gradients that should go where the bands are, and
8628 It is designed for playback only. Do not use it prior to
8629 lossy compression, because compression tends to lose the dither and
8630 bring back the bands.
8632 It accepts the following parameters:
8637 The maximum amount by which the filter will change any one pixel. This is also
8638 the threshold for detecting nearly flat regions. Acceptable values range from
8639 .51 to 64; the default value is 1.2. Out-of-range values will be clipped to the
8643 The neighborhood to fit the gradient to. A larger radius makes for smoother
8644 gradients, but also prevents the filter from modifying the pixels near detailed
8645 regions. Acceptable values are 8-32; the default value is 16. Out-of-range
8646 values will be clipped to the valid range.
8650 Alternatively, the options can be specified as a flat string:
8651 @var{strength}[:@var{radius}]
8653 @subsection Examples
8657 Apply the filter with a @code{3.5} strength and radius of @code{8}:
8663 Specify radius, omitting the strength (which will fall-back to the default
8674 Apply a Hald CLUT to a video stream.
8676 First input is the video stream to process, and second one is the Hald CLUT.
8677 The Hald CLUT input can be a simple picture or a complete video stream.
8679 The filter accepts the following options:
8683 Force termination when the shortest input terminates. Default is @code{0}.
8685 Continue applying the last CLUT after the end of the stream. A value of
8686 @code{0} disable the filter after the last frame of the CLUT is reached.
8687 Default is @code{1}.
8690 @code{haldclut} also has the same interpolation options as @ref{lut3d} (both
8691 filters share the same internals).
8693 More information about the Hald CLUT can be found on Eskil Steenberg's website
8694 (Hald CLUT author) at @url{http://www.quelsolaar.com/technology/clut.html}.
8696 @subsection Workflow examples
8698 @subsubsection Hald CLUT video stream
8700 Generate an identity Hald CLUT stream altered with various effects:
8702 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "hue=H=2*PI*t:s=sin(2*PI*t)+1, curves=cross_process" -t 10 -c:v ffv1 clut.nut
8705 Note: make sure you use a lossless codec.
8707 Then use it with @code{haldclut} to apply it on some random stream:
8709 ffmpeg -f lavfi -i mandelbrot -i clut.nut -filter_complex '[0][1] haldclut' -t 20 mandelclut.mkv
8712 The Hald CLUT will be applied to the 10 first seconds (duration of
8713 @file{clut.nut}), then the latest picture of that CLUT stream will be applied
8714 to the remaining frames of the @code{mandelbrot} stream.
8716 @subsubsection Hald CLUT with preview
8718 A Hald CLUT is supposed to be a squared image of @code{Level*Level*Level} by
8719 @code{Level*Level*Level} pixels. For a given Hald CLUT, FFmpeg will select the
8720 biggest possible square starting at the top left of the picture. The remaining
8721 padding pixels (bottom or right) will be ignored. This area can be used to add
8722 a preview of the Hald CLUT.
8724 Typically, the following generated Hald CLUT will be supported by the
8725 @code{haldclut} filter:
8728 ffmpeg -f lavfi -i @ref{haldclutsrc}=8 -vf "
8729 pad=iw+320 [padded_clut];
8730 smptebars=s=320x256, split [a][b];
8731 [padded_clut][a] overlay=W-320:h, curves=color_negative [main];
8732 [main][b] overlay=W-320" -frames:v 1 clut.png
8735 It contains the original and a preview of the effect of the CLUT: SMPTE color
8736 bars are displayed on the right-top, and below the same color bars processed by
8739 Then, the effect of this Hald CLUT can be visualized with:
8741 ffplay input.mkv -vf "movie=clut.png, [in] haldclut"
8746 Flip the input video horizontally.
8748 For example, to horizontally flip the input video with @command{ffmpeg}:
8750 ffmpeg -i in.avi -vf "hflip" out.avi
8754 This filter applies a global color histogram equalization on a
8757 It can be used to correct video that has a compressed range of pixel
8758 intensities. The filter redistributes the pixel intensities to
8759 equalize their distribution across the intensity range. It may be
8760 viewed as an "automatically adjusting contrast filter". This filter is
8761 useful only for correcting degraded or poorly captured source
8764 The filter accepts the following options:
8768 Determine the amount of equalization to be applied. As the strength
8769 is reduced, the distribution of pixel intensities more-and-more
8770 approaches that of the input frame. The value must be a float number
8771 in the range [0,1] and defaults to 0.200.
8774 Set the maximum intensity that can generated and scale the output
8775 values appropriately. The strength should be set as desired and then
8776 the intensity can be limited if needed to avoid washing-out. The value
8777 must be a float number in the range [0,1] and defaults to 0.210.
8780 Set the antibanding level. If enabled the filter will randomly vary
8781 the luminance of output pixels by a small amount to avoid banding of
8782 the histogram. Possible values are @code{none}, @code{weak} or
8783 @code{strong}. It defaults to @code{none}.
8788 Compute and draw a color distribution histogram for the input video.
8790 The computed histogram is a representation of the color component
8791 distribution in an image.
8793 Standard histogram displays the color components distribution in an image.
8794 Displays color graph for each color component. Shows distribution of
8795 the Y, U, V, A or R, G, B components, depending on input format, in the
8796 current frame. Below each graph a color component scale meter is shown.
8798 The filter accepts the following options:
8802 Set height of level. Default value is @code{200}.
8803 Allowed range is [50, 2048].
8806 Set height of color scale. Default value is @code{12}.
8807 Allowed range is [0, 40].
8811 It accepts the following values:
8814 Per color component graphs are placed below each other.
8817 Per color component graphs are placed side by side.
8820 Presents information identical to that in the @code{parade}, except
8821 that the graphs representing color components are superimposed directly
8824 Default is @code{stack}.
8827 Set mode. Can be either @code{linear}, or @code{logarithmic}.
8828 Default is @code{linear}.
8831 Set what color components to display.
8832 Default is @code{7}.
8835 Set foreground opacity. Default is @code{0.7}.
8838 Set background opacity. Default is @code{0.5}.
8841 @subsection Examples
8846 Calculate and draw histogram:
8848 ffplay -i input -vf histogram
8856 This is a high precision/quality 3d denoise filter. It aims to reduce
8857 image noise, producing smooth images and making still images really
8858 still. It should enhance compressibility.
8860 It accepts the following optional parameters:
8864 A non-negative floating point number which specifies spatial luma strength.
8867 @item chroma_spatial
8868 A non-negative floating point number which specifies spatial chroma strength.
8869 It defaults to 3.0*@var{luma_spatial}/4.0.
8872 A floating point number which specifies luma temporal strength. It defaults to
8873 6.0*@var{luma_spatial}/4.0.
8876 A floating point number which specifies chroma temporal strength. It defaults to
8877 @var{luma_tmp}*@var{chroma_spatial}/@var{luma_spatial}.
8880 @anchor{hwupload_cuda}
8881 @section hwupload_cuda
8883 Upload system memory frames to a CUDA device.
8885 It accepts the following optional parameters:
8889 The number of the CUDA device to use
8894 Apply a high-quality magnification filter designed for pixel art. This filter
8895 was originally created by Maxim Stepin.
8897 It accepts the following option:
8901 Set the scaling dimension: @code{2} for @code{hq2x}, @code{3} for
8902 @code{hq3x} and @code{4} for @code{hq4x}.
8903 Default is @code{3}.
8907 Stack input videos horizontally.
8909 All streams must be of same pixel format and of same height.
8911 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
8912 to create same output.
8914 The filter accept the following option:
8918 Set number of input streams. Default is 2.
8921 If set to 1, force the output to terminate when the shortest input
8922 terminates. Default value is 0.
8927 Modify the hue and/or the saturation of the input.
8929 It accepts the following parameters:
8933 Specify the hue angle as a number of degrees. It accepts an expression,
8934 and defaults to "0".
8937 Specify the saturation in the [-10,10] range. It accepts an expression and
8941 Specify the hue angle as a number of radians. It accepts an
8942 expression, and defaults to "0".
8945 Specify the brightness in the [-10,10] range. It accepts an expression and
8949 @option{h} and @option{H} are mutually exclusive, and can't be
8950 specified at the same time.
8952 The @option{b}, @option{h}, @option{H} and @option{s} option values are
8953 expressions containing the following constants:
8957 frame count of the input frame starting from 0
8960 presentation timestamp of the input frame expressed in time base units
8963 frame rate of the input video, NAN if the input frame rate is unknown
8966 timestamp expressed in seconds, NAN if the input timestamp is unknown
8969 time base of the input video
8972 @subsection Examples
8976 Set the hue to 90 degrees and the saturation to 1.0:
8982 Same command but expressing the hue in radians:
8988 Rotate hue and make the saturation swing between 0
8989 and 2 over a period of 1 second:
8991 hue="H=2*PI*t: s=sin(2*PI*t)+1"
8995 Apply a 3 seconds saturation fade-in effect starting at 0:
9000 The general fade-in expression can be written as:
9002 hue="s=min(0\, max((t-START)/DURATION\, 1))"
9006 Apply a 3 seconds saturation fade-out effect starting at 5 seconds:
9008 hue="s=max(0\, min(1\, (8-t)/3))"
9011 The general fade-out expression can be written as:
9013 hue="s=max(0\, min(1\, (START+DURATION-t)/DURATION))"
9018 @subsection Commands
9020 This filter supports the following commands:
9026 Modify the hue and/or the saturation and/or brightness of the input video.
9027 The command accepts the same syntax of the corresponding option.
9029 If the specified expression is not valid, it is kept at its current
9035 Grow first stream into second stream by connecting components.
9036 This makes it possible to build more robust edge masks.
9038 This filter accepts the following options:
9042 Set which planes will be processed as bitmap, unprocessed planes will be
9043 copied from first stream.
9044 By default value 0xf, all planes will be processed.
9047 Set threshold which is used in filtering. If pixel component value is higher than
9048 this value filter algorithm for connecting components is activated.
9049 By default value is 0.
9054 Detect video interlacing type.
9056 This filter tries to detect if the input frames are interlaced, progressive,
9057 top or bottom field first. It will also try to detect fields that are
9058 repeated between adjacent frames (a sign of telecine).
9060 Single frame detection considers only immediately adjacent frames when classifying each frame.
9061 Multiple frame detection incorporates the classification history of previous frames.
9063 The filter will log these metadata values:
9066 @item single.current_frame
9067 Detected type of current frame using single-frame detection. One of:
9068 ``tff'' (top field first), ``bff'' (bottom field first),
9069 ``progressive'', or ``undetermined''
9072 Cumulative number of frames detected as top field first using single-frame detection.
9075 Cumulative number of frames detected as top field first using multiple-frame detection.
9078 Cumulative number of frames detected as bottom field first using single-frame detection.
9080 @item multiple.current_frame
9081 Detected type of current frame using multiple-frame detection. One of:
9082 ``tff'' (top field first), ``bff'' (bottom field first),
9083 ``progressive'', or ``undetermined''
9086 Cumulative number of frames detected as bottom field first using multiple-frame detection.
9088 @item single.progressive
9089 Cumulative number of frames detected as progressive using single-frame detection.
9091 @item multiple.progressive
9092 Cumulative number of frames detected as progressive using multiple-frame detection.
9094 @item single.undetermined
9095 Cumulative number of frames that could not be classified using single-frame detection.
9097 @item multiple.undetermined
9098 Cumulative number of frames that could not be classified using multiple-frame detection.
9100 @item repeated.current_frame
9101 Which field in the current frame is repeated from the last. One of ``neither'', ``top'', or ``bottom''.
9103 @item repeated.neither
9104 Cumulative number of frames with no repeated field.
9107 Cumulative number of frames with the top field repeated from the previous frame's top field.
9109 @item repeated.bottom
9110 Cumulative number of frames with the bottom field repeated from the previous frame's bottom field.
9113 The filter accepts the following options:
9117 Set interlacing threshold.
9119 Set progressive threshold.
9121 Threshold for repeated field detection.
9123 Number of frames after which a given frame's contribution to the
9124 statistics is halved (i.e., it contributes only 0.5 to its
9125 classification). The default of 0 means that all frames seen are given
9126 full weight of 1.0 forever.
9127 @item analyze_interlaced_flag
9128 When this is not 0 then idet will use the specified number of frames to determine
9129 if the interlaced flag is accurate, it will not count undetermined frames.
9130 If the flag is found to be accurate it will be used without any further
9131 computations, if it is found to be inaccurate it will be cleared without any
9132 further computations. This allows inserting the idet filter as a low computational
9133 method to clean up the interlaced flag
9138 Deinterleave or interleave fields.
9140 This filter allows one to process interlaced images fields without
9141 deinterlacing them. Deinterleaving splits the input frame into 2
9142 fields (so called half pictures). Odd lines are moved to the top
9143 half of the output image, even lines to the bottom half.
9144 You can process (filter) them independently and then re-interleave them.
9146 The filter accepts the following options:
9150 @item chroma_mode, c
9152 Available values for @var{luma_mode}, @var{chroma_mode} and
9153 @var{alpha_mode} are:
9159 @item deinterleave, d
9160 Deinterleave fields, placing one above the other.
9163 Interleave fields. Reverse the effect of deinterleaving.
9165 Default value is @code{none}.
9168 @item chroma_swap, cs
9169 @item alpha_swap, as
9170 Swap luma/chroma/alpha fields. Exchange even & odd lines. Default value is @code{0}.
9175 Apply inflate effect to the video.
9177 This filter replaces the pixel by the local(3x3) average by taking into account
9178 only values higher than the pixel.
9180 It accepts the following options:
9187 Limit the maximum change for each plane, default is 65535.
9188 If 0, plane will remain unchanged.
9193 Simple interlacing filter from progressive contents. This interleaves upper (or
9194 lower) lines from odd frames with lower (or upper) lines from even frames,
9195 halving the frame rate and preserving image height.
9198 Original Original New Frame
9199 Frame 'j' Frame 'j+1' (tff)
9200 ========== =========== ==================
9201 Line 0 --------------------> Frame 'j' Line 0
9202 Line 1 Line 1 ----> Frame 'j+1' Line 1
9203 Line 2 ---------------------> Frame 'j' Line 2
9204 Line 3 Line 3 ----> Frame 'j+1' Line 3
9206 New Frame + 1 will be generated by Frame 'j+2' and Frame 'j+3' and so on
9209 It accepts the following optional parameters:
9213 This determines whether the interlaced frame is taken from the even
9214 (tff - default) or odd (bff) lines of the progressive frame.
9217 Vertical lowpass filter to avoid twitter interlacing and
9218 reduce moire patterns.
9222 Disable vertical lowpass filter
9225 Enable linear filter (default)
9228 Enable complex filter. This will slightly less reduce twitter and moire
9229 but better retain detail and subjective sharpness impression.
9236 Deinterlace input video by applying Donald Graft's adaptive kernel
9237 deinterling. Work on interlaced parts of a video to produce
9240 The description of the accepted parameters follows.
9244 Set the threshold which affects the filter's tolerance when
9245 determining if a pixel line must be processed. It must be an integer
9246 in the range [0,255] and defaults to 10. A value of 0 will result in
9247 applying the process on every pixels.
9250 Paint pixels exceeding the threshold value to white if set to 1.
9254 Set the fields order. Swap fields if set to 1, leave fields alone if
9258 Enable additional sharpening if set to 1. Default is 0.
9261 Enable twoway sharpening if set to 1. Default is 0.
9264 @subsection Examples
9268 Apply default values:
9270 kerndeint=thresh=10:map=0:order=0:sharp=0:twoway=0
9274 Enable additional sharpening:
9280 Paint processed pixels in white:
9286 @section lenscorrection
9288 Correct radial lens distortion
9290 This filter can be used to correct for radial distortion as can result from the use
9291 of wide angle lenses, and thereby re-rectify the image. To find the right parameters
9292 one can use tools available for example as part of opencv or simply trial-and-error.
9293 To use opencv use the calibration sample (under samples/cpp) from the opencv sources
9294 and extract the k1 and k2 coefficients from the resulting matrix.
9296 Note that effectively the same filter is available in the open-source tools Krita and
9297 Digikam from the KDE project.
9299 In contrast to the @ref{vignette} filter, which can also be used to compensate lens errors,
9300 this filter corrects the distortion of the image, whereas @ref{vignette} corrects the
9301 brightness distribution, so you may want to use both filters together in certain
9302 cases, though you will have to take care of ordering, i.e. whether vignetting should
9303 be applied before or after lens correction.
9307 The filter accepts the following options:
9311 Relative x-coordinate of the focal point of the image, and thereby the center of the
9312 distortion. This value has a range [0,1] and is expressed as fractions of the image
9315 Relative y-coordinate of the focal point of the image, and thereby the center of the
9316 distortion. This value has a range [0,1] and is expressed as fractions of the image
9319 Coefficient of the quadratic correction term. 0.5 means no correction.
9321 Coefficient of the double quadratic correction term. 0.5 means no correction.
9324 The formula that generates the correction is:
9326 @var{r_src} = @var{r_tgt} * (1 + @var{k1} * (@var{r_tgt} / @var{r_0})^2 + @var{k2} * (@var{r_tgt} / @var{r_0})^4)
9328 where @var{r_0} is halve of the image diagonal and @var{r_src} and @var{r_tgt} are the
9329 distances from the focal point in the source and target images, respectively.
9335 The filter accepts the following options:
9339 Set the number of loops.
9342 Set maximal size in number of frames.
9345 Set first frame of loop.
9351 Apply a 3D LUT to an input video.
9353 The filter accepts the following options:
9357 Set the 3D LUT file name.
9359 Currently supported formats:
9371 Select interpolation mode.
9373 Available values are:
9377 Use values from the nearest defined point.
9379 Interpolate values using the 8 points defining a cube.
9381 Interpolate values using a tetrahedron.
9387 Turn certain luma values into transparency.
9389 The filter accepts the following options:
9393 Set the luma which will be used as base for transparency.
9394 Default value is @code{0}.
9397 Set the range of luma values to be keyed out.
9398 Default value is @code{0}.
9401 Set the range of softness. Default value is @code{0}.
9402 Use this to control gradual transition from zero to full transparency.
9405 @section lut, lutrgb, lutyuv
9407 Compute a look-up table for binding each pixel component input value
9408 to an output value, and apply it to the input video.
9410 @var{lutyuv} applies a lookup table to a YUV input video, @var{lutrgb}
9411 to an RGB input video.
9413 These filters accept the following parameters:
9416 set first pixel component expression
9418 set second pixel component expression
9420 set third pixel component expression
9422 set fourth pixel component expression, corresponds to the alpha component
9425 set red component expression
9427 set green component expression
9429 set blue component expression
9431 alpha component expression
9434 set Y/luminance component expression
9436 set U/Cb component expression
9438 set V/Cr component expression
9441 Each of them specifies the expression to use for computing the lookup table for
9442 the corresponding pixel component values.
9444 The exact component associated to each of the @var{c*} options depends on the
9447 The @var{lut} filter requires either YUV or RGB pixel formats in input,
9448 @var{lutrgb} requires RGB pixel formats in input, and @var{lutyuv} requires YUV.
9450 The expressions can contain the following constants and functions:
9455 The input width and height.
9458 The input value for the pixel component.
9461 The input value, clipped to the @var{minval}-@var{maxval} range.
9464 The maximum value for the pixel component.
9467 The minimum value for the pixel component.
9470 The negated value for the pixel component value, clipped to the
9471 @var{minval}-@var{maxval} range; it corresponds to the expression
9472 "maxval-clipval+minval".
9475 The computed value in @var{val}, clipped to the
9476 @var{minval}-@var{maxval} range.
9478 @item gammaval(gamma)
9479 The computed gamma correction value of the pixel component value,
9480 clipped to the @var{minval}-@var{maxval} range. It corresponds to the
9482 "pow((clipval-minval)/(maxval-minval)\,@var{gamma})*(maxval-minval)+minval"
9486 All expressions default to "val".
9488 @subsection Examples
9494 lutrgb="r=maxval+minval-val:g=maxval+minval-val:b=maxval+minval-val"
9495 lutyuv="y=maxval+minval-val:u=maxval+minval-val:v=maxval+minval-val"
9498 The above is the same as:
9500 lutrgb="r=negval:g=negval:b=negval"
9501 lutyuv="y=negval:u=negval:v=negval"
9511 Remove chroma components, turning the video into a graytone image:
9513 lutyuv="u=128:v=128"
9517 Apply a luma burning effect:
9523 Remove green and blue components:
9529 Set a constant alpha channel value on input:
9531 format=rgba,lutrgb=a="maxval-minval/2"
9535 Correct luminance gamma by a factor of 0.5:
9537 lutyuv=y=gammaval(0.5)
9541 Discard least significant bits of luma:
9543 lutyuv=y='bitand(val, 128+64+32)'
9547 Technicolor like effect:
9549 lutyuv=u='(val-maxval/2)*2+maxval/2':v='(val-maxval/2)*2+maxval/2'
9555 Compute and apply a lookup table from two video inputs.
9557 This filter accepts the following parameters:
9560 set first pixel component expression
9562 set second pixel component expression
9564 set third pixel component expression
9566 set fourth pixel component expression, corresponds to the alpha component
9569 Each of them specifies the expression to use for computing the lookup table for
9570 the corresponding pixel component values.
9572 The exact component associated to each of the @var{c*} options depends on the
9575 The expressions can contain the following constants:
9580 The input width and height.
9583 The first input value for the pixel component.
9586 The second input value for the pixel component.
9589 The first input video bit depth.
9592 The second input video bit depth.
9595 All expressions default to "x".
9597 @subsection Examples
9601 Highlight differences between two RGB video streams:
9603 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,0,pow(2,bdx)-1)'
9607 Highlight differences between two YUV video streams:
9609 lut2='ifnot(x-y,0,pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1):ifnot(x-y,pow(2,bdx-1),pow(2,bdx)-1)'
9613 @section maskedclamp
9615 Clamp the first input stream with the second input and third input stream.
9617 Returns the value of first stream to be between second input
9618 stream - @code{undershoot} and third input stream + @code{overshoot}.
9620 This filter accepts the following options:
9623 Default value is @code{0}.
9626 Default value is @code{0}.
9629 Set which planes will be processed as bitmap, unprocessed planes will be
9630 copied from first stream.
9631 By default value 0xf, all planes will be processed.
9634 @section maskedmerge
9636 Merge the first input stream with the second input stream using per pixel
9637 weights in the third input stream.
9639 A value of 0 in the third stream pixel component means that pixel component
9640 from first stream is returned unchanged, while maximum value (eg. 255 for
9641 8-bit videos) means that pixel component from second stream is returned
9642 unchanged. Intermediate values define the amount of merging between both
9643 input stream's pixel components.
9645 This filter accepts the following options:
9648 Set which planes will be processed as bitmap, unprocessed planes will be
9649 copied from first stream.
9650 By default value 0xf, all planes will be processed.
9655 Apply motion-compensation deinterlacing.
9657 It needs one field per frame as input and must thus be used together
9658 with yadif=1/3 or equivalent.
9660 This filter accepts the following options:
9663 Set the deinterlacing mode.
9665 It accepts one of the following values:
9670 use iterative motion estimation
9672 like @samp{slow}, but use multiple reference frames.
9674 Default value is @samp{fast}.
9677 Set the picture field parity assumed for the input video. It must be
9678 one of the following values:
9682 assume top field first
9684 assume bottom field first
9687 Default value is @samp{bff}.
9690 Set per-block quantization parameter (QP) used by the internal
9693 Higher values should result in a smoother motion vector field but less
9694 optimal individual vectors. Default value is 1.
9697 @section mergeplanes
9699 Merge color channel components from several video streams.
9701 The filter accepts up to 4 input streams, and merge selected input
9702 planes to the output video.
9704 This filter accepts the following options:
9707 Set input to output plane mapping. Default is @code{0}.
9709 The mappings is specified as a bitmap. It should be specified as a
9710 hexadecimal number in the form 0xAa[Bb[Cc[Dd]]]. 'Aa' describes the
9711 mapping for the first plane of the output stream. 'A' sets the number of
9712 the input stream to use (from 0 to 3), and 'a' the plane number of the
9713 corresponding input to use (from 0 to 3). The rest of the mappings is
9714 similar, 'Bb' describes the mapping for the output stream second
9715 plane, 'Cc' describes the mapping for the output stream third plane and
9716 'Dd' describes the mapping for the output stream fourth plane.
9719 Set output pixel format. Default is @code{yuva444p}.
9722 @subsection Examples
9726 Merge three gray video streams of same width and height into single video stream:
9728 [a0][a1][a2]mergeplanes=0x001020:yuv444p
9732 Merge 1st yuv444p stream and 2nd gray video stream into yuva444p video stream:
9734 [a0][a1]mergeplanes=0x00010210:yuva444p
9738 Swap Y and A plane in yuva444p stream:
9740 format=yuva444p,mergeplanes=0x03010200:yuva444p
9744 Swap U and V plane in yuv420p stream:
9746 format=yuv420p,mergeplanes=0x000201:yuv420p
9750 Cast a rgb24 clip to yuv444p:
9752 format=rgb24,mergeplanes=0x000102:yuv444p
9758 Estimate and export motion vectors using block matching algorithms.
9759 Motion vectors are stored in frame side data to be used by other filters.
9761 This filter accepts the following options:
9764 Specify the motion estimation method. Accepts one of the following values:
9768 Exhaustive search algorithm.
9770 Three step search algorithm.
9772 Two dimensional logarithmic search algorithm.
9774 New three step search algorithm.
9776 Four step search algorithm.
9778 Diamond search algorithm.
9780 Hexagon-based search algorithm.
9782 Enhanced predictive zonal search algorithm.
9784 Uneven multi-hexagon search algorithm.
9786 Default value is @samp{esa}.
9789 Macroblock size. Default @code{16}.
9792 Search parameter. Default @code{7}.
9795 @section midequalizer
9797 Apply Midway Image Equalization effect using two video streams.
9799 Midway Image Equalization adjusts a pair of images to have the same
9800 histogram, while maintaining their dynamics as much as possible. It's
9801 useful for e.g. matching exposures from a pair of stereo cameras.
9803 This filter has two inputs and one output, which must be of same pixel format, but
9804 may be of different sizes. The output of filter is first input adjusted with
9805 midway histogram of both inputs.
9807 This filter accepts the following option:
9811 Set which planes to process. Default is @code{15}, which is all available planes.
9814 @section minterpolate
9816 Convert the video to specified frame rate using motion interpolation.
9818 This filter accepts the following options:
9821 Specify the output frame rate. This can be rational e.g. @code{60000/1001}. Frames are dropped if @var{fps} is lower than source fps. Default @code{60}.
9824 Motion interpolation mode. Following values are accepted:
9827 Duplicate previous or next frame for interpolating new ones.
9829 Blend source frames. Interpolated frame is mean of previous and next frames.
9831 Motion compensated interpolation. Following options are effective when this mode is selected:
9835 Motion compensation mode. Following values are accepted:
9838 Overlapped block motion compensation.
9840 Adaptive overlapped block motion compensation. Window weighting coefficients are controlled adaptively according to the reliabilities of the neighboring motion vectors to reduce oversmoothing.
9842 Default mode is @samp{obmc}.
9845 Motion estimation mode. Following values are accepted:
9848 Bidirectional motion estimation. Motion vectors are estimated for each source frame in both forward and backward directions.
9850 Bilateral motion estimation. Motion vectors are estimated directly for interpolated frame.
9852 Default mode is @samp{bilat}.
9855 The algorithm to be used for motion estimation. Following values are accepted:
9858 Exhaustive search algorithm.
9860 Three step search algorithm.
9862 Two dimensional logarithmic search algorithm.
9864 New three step search algorithm.
9866 Four step search algorithm.
9868 Diamond search algorithm.
9870 Hexagon-based search algorithm.
9872 Enhanced predictive zonal search algorithm.
9874 Uneven multi-hexagon search algorithm.
9876 Default algorithm is @samp{epzs}.
9879 Macroblock size. Default @code{16}.
9882 Motion estimation search parameter. Default @code{32}.
9885 Enable variable-size block motion compensation. Motion estimation is applied with smaller block sizes at object boundaries in order to make the them less blur. Default is @code{0} (disabled).
9890 Scene change detection method. Scene change leads motion vectors to be in random direction. Scene change detection replace interpolated frames by duplicate ones. May not be needed for other modes. Following values are accepted:
9893 Disable scene change detection.
9895 Frame difference. Corresponding pixel values are compared and if it satisfies @var{scd_threshold} scene change is detected.
9897 Default method is @samp{fdiff}.
9900 Scene change detection threshold. Default is @code{5.0}.
9905 Drop frames that do not differ greatly from the previous frame in
9906 order to reduce frame rate.
9908 The main use of this filter is for very-low-bitrate encoding
9909 (e.g. streaming over dialup modem), but it could in theory be used for
9910 fixing movies that were inverse-telecined incorrectly.
9912 A description of the accepted options follows.
9916 Set the maximum number of consecutive frames which can be dropped (if
9917 positive), or the minimum interval between dropped frames (if
9918 negative). If the value is 0, the frame is dropped unregarding the
9919 number of previous sequentially dropped frames.
9926 Set the dropping threshold values.
9928 Values for @option{hi} and @option{lo} are for 8x8 pixel blocks and
9929 represent actual pixel value differences, so a threshold of 64
9930 corresponds to 1 unit of difference for each pixel, or the same spread
9931 out differently over the block.
9933 A frame is a candidate for dropping if no 8x8 blocks differ by more
9934 than a threshold of @option{hi}, and if no more than @option{frac} blocks (1
9935 meaning the whole image) differ by more than a threshold of @option{lo}.
9937 Default value for @option{hi} is 64*12, default value for @option{lo} is
9938 64*5, and default value for @option{frac} is 0.33.
9946 It accepts an integer in input; if non-zero it negates the
9947 alpha component (if available). The default value in input is 0.
9951 Denoise frames using Non-Local Means algorithm.
9953 Each pixel is adjusted by looking for other pixels with similar contexts. This
9954 context similarity is defined by comparing their surrounding patches of size
9955 @option{p}x@option{p}. Patches are searched in an area of @option{r}x@option{r}
9958 Note that the research area defines centers for patches, which means some
9959 patches will be made of pixels outside that research area.
9961 The filter accepts the following options.
9965 Set denoising strength.
9971 Same as @option{p} but for chroma planes.
9973 The default value is @var{0} and means automatic.
9979 Same as @option{r} but for chroma planes.
9981 The default value is @var{0} and means automatic.
9986 Deinterlace video using neural network edge directed interpolation.
9988 This filter accepts the following options:
9992 Mandatory option, without binary file filter can not work.
9993 Currently file can be found here:
9994 https://github.com/dubhater/vapoursynth-nnedi3/blob/master/src/nnedi3_weights.bin
9997 Set which frames to deinterlace, by default it is @code{all}.
9998 Can be @code{all} or @code{interlaced}.
10001 Set mode of operation.
10003 Can be one of the following:
10007 Use frame flags, both fields.
10009 Use frame flags, single field.
10011 Use top field only.
10013 Use bottom field only.
10015 Use both fields, top first.
10017 Use both fields, bottom first.
10021 Set which planes to process, by default filter process all frames.
10024 Set size of local neighborhood around each pixel, used by the predictor neural
10027 Can be one of the following:
10040 Set the number of neurons in predicctor neural network.
10041 Can be one of the following:
10052 Controls the number of different neural network predictions that are blended
10053 together to compute the final output value. Can be @code{fast}, default or
10057 Set which set of weights to use in the predictor.
10058 Can be one of the following:
10062 weights trained to minimize absolute error
10064 weights trained to minimize squared error
10068 Controls whether or not the prescreener neural network is used to decide
10069 which pixels should be processed by the predictor neural network and which
10070 can be handled by simple cubic interpolation.
10071 The prescreener is trained to know whether cubic interpolation will be
10072 sufficient for a pixel or whether it should be predicted by the predictor nn.
10073 The computational complexity of the prescreener nn is much less than that of
10074 the predictor nn. Since most pixels can be handled by cubic interpolation,
10075 using the prescreener generally results in much faster processing.
10076 The prescreener is pretty accurate, so the difference between using it and not
10077 using it is almost always unnoticeable.
10079 Can be one of the following:
10087 Default is @code{new}.
10090 Set various debugging flags.
10095 Force libavfilter not to use any of the specified pixel formats for the
10096 input to the next filter.
10098 It accepts the following parameters:
10102 A '|'-separated list of pixel format names, such as
10103 apix_fmts=yuv420p|monow|rgb24".
10107 @subsection Examples
10111 Force libavfilter to use a format different from @var{yuv420p} for the
10112 input to the vflip filter:
10114 noformat=pix_fmts=yuv420p,vflip
10118 Convert the input video to any of the formats not contained in the list:
10120 noformat=yuv420p|yuv444p|yuv410p
10126 Add noise on video input frame.
10128 The filter accepts the following options:
10136 Set noise seed for specific pixel component or all pixel components in case
10137 of @var{all_seed}. Default value is @code{123457}.
10139 @item all_strength, alls
10140 @item c0_strength, c0s
10141 @item c1_strength, c1s
10142 @item c2_strength, c2s
10143 @item c3_strength, c3s
10144 Set noise strength for specific pixel component or all pixel components in case
10145 @var{all_strength}. Default value is @code{0}. Allowed range is [0, 100].
10147 @item all_flags, allf
10148 @item c0_flags, c0f
10149 @item c1_flags, c1f
10150 @item c2_flags, c2f
10151 @item c3_flags, c3f
10152 Set pixel component flags or set flags for all components if @var{all_flags}.
10153 Available values for component flags are:
10156 averaged temporal noise (smoother)
10158 mix random noise with a (semi)regular pattern
10160 temporal noise (noise pattern changes between frames)
10162 uniform noise (gaussian otherwise)
10166 @subsection Examples
10168 Add temporal and uniform noise to input video:
10170 noise=alls=20:allf=t+u
10175 Pass the video source unchanged to the output.
10178 Optical Character Recognition
10180 This filter uses Tesseract for optical character recognition.
10182 It accepts the following options:
10186 Set datapath to tesseract data. Default is to use whatever was
10187 set at installation.
10190 Set language, default is "eng".
10193 Set character whitelist.
10196 Set character blacklist.
10199 The filter exports recognized text as the frame metadata @code{lavfi.ocr.text}.
10203 Apply a video transform using libopencv.
10205 To enable this filter, install the libopencv library and headers and
10206 configure FFmpeg with @code{--enable-libopencv}.
10208 It accepts the following parameters:
10213 The name of the libopencv filter to apply.
10215 @item filter_params
10216 The parameters to pass to the libopencv filter. If not specified, the default
10217 values are assumed.
10221 Refer to the official libopencv documentation for more precise
10223 @url{http://docs.opencv.org/master/modules/imgproc/doc/filtering.html}
10225 Several libopencv filters are supported; see the following subsections.
10230 Dilate an image by using a specific structuring element.
10231 It corresponds to the libopencv function @code{cvDilate}.
10233 It accepts the parameters: @var{struct_el}|@var{nb_iterations}.
10235 @var{struct_el} represents a structuring element, and has the syntax:
10236 @var{cols}x@var{rows}+@var{anchor_x}x@var{anchor_y}/@var{shape}
10238 @var{cols} and @var{rows} represent the number of columns and rows of
10239 the structuring element, @var{anchor_x} and @var{anchor_y} the anchor
10240 point, and @var{shape} the shape for the structuring element. @var{shape}
10241 must be "rect", "cross", "ellipse", or "custom".
10243 If the value for @var{shape} is "custom", it must be followed by a
10244 string of the form "=@var{filename}". The file with name
10245 @var{filename} is assumed to represent a binary image, with each
10246 printable character corresponding to a bright pixel. When a custom
10247 @var{shape} is used, @var{cols} and @var{rows} are ignored, the number
10248 or columns and rows of the read file are assumed instead.
10250 The default value for @var{struct_el} is "3x3+0x0/rect".
10252 @var{nb_iterations} specifies the number of times the transform is
10253 applied to the image, and defaults to 1.
10257 # Use the default values
10260 # Dilate using a structuring element with a 5x5 cross, iterating two times
10261 ocv=filter_name=dilate:filter_params=5x5+2x2/cross|2
10263 # Read the shape from the file diamond.shape, iterating two times.
10264 # The file diamond.shape may contain a pattern of characters like this
10270 # The specified columns and rows are ignored
10271 # but the anchor point coordinates are not
10272 ocv=dilate:0x0+2x2/custom=diamond.shape|2
10277 Erode an image by using a specific structuring element.
10278 It corresponds to the libopencv function @code{cvErode}.
10280 It accepts the parameters: @var{struct_el}:@var{nb_iterations},
10281 with the same syntax and semantics as the @ref{dilate} filter.
10285 Smooth the input video.
10287 The filter takes the following parameters:
10288 @var{type}|@var{param1}|@var{param2}|@var{param3}|@var{param4}.
10290 @var{type} is the type of smooth filter to apply, and must be one of
10291 the following values: "blur", "blur_no_scale", "median", "gaussian",
10292 or "bilateral". The default value is "gaussian".
10294 The meaning of @var{param1}, @var{param2}, @var{param3}, and @var{param4}
10295 depend on the smooth type. @var{param1} and
10296 @var{param2} accept integer positive values or 0. @var{param3} and
10297 @var{param4} accept floating point values.
10299 The default value for @var{param1} is 3. The default value for the
10300 other parameters is 0.
10302 These parameters correspond to the parameters assigned to the
10303 libopencv function @code{cvSmooth}.
10305 @section oscilloscope
10307 2D Video Oscilloscope.
10309 Useful to measure spatial impulse, step responses, chroma delays, etc.
10311 It accepts the following parameters:
10315 Set scope center x position.
10318 Set scope center y position.
10321 Set scope size, relative to frame diagonal.
10324 Set scope tilt/rotation.
10330 Set trace center x position.
10333 Set trace center y position.
10336 Set trace width, relative to width of frame.
10339 Set trace height, relative to height of frame.
10342 Set which components to trace. By default it traces first three components.
10345 Draw trace grid. By default is enabled.
10348 Draw some statistics. By default is enabled.
10351 Draw scope. By default is enabled.
10354 @subsection Examples
10358 Inspect full first row of video frame.
10360 oscilloscope=x=0.5:y=0:s=1
10364 Inspect full last row of video frame.
10366 oscilloscope=x=0.5:y=1:s=1
10370 Inspect full 5th line of video frame of height 1080.
10372 oscilloscope=x=0.5:y=5/1080:s=1
10376 Inspect full last column of video frame.
10378 oscilloscope=x=1:y=0.5:s=1:t=1
10386 Overlay one video on top of another.
10388 It takes two inputs and has one output. The first input is the "main"
10389 video on which the second input is overlaid.
10391 It accepts the following parameters:
10393 A description of the accepted options follows.
10398 Set the expression for the x and y coordinates of the overlaid video
10399 on the main video. Default value is "0" for both expressions. In case
10400 the expression is invalid, it is set to a huge value (meaning that the
10401 overlay will not be displayed within the output visible area).
10404 The action to take when EOF is encountered on the secondary input; it accepts
10405 one of the following values:
10409 Repeat the last frame (the default).
10413 Pass the main input through.
10417 Set when the expressions for @option{x}, and @option{y} are evaluated.
10419 It accepts the following values:
10422 only evaluate expressions once during the filter initialization or
10423 when a command is processed
10426 evaluate expressions for each incoming frame
10429 Default value is @samp{frame}.
10432 If set to 1, force the output to terminate when the shortest input
10433 terminates. Default value is 0.
10436 Set the format for the output video.
10438 It accepts the following values:
10441 force YUV420 output
10444 force YUV422 output
10447 force YUV444 output
10450 force packed RGB output
10453 force planar RGB output
10456 Default value is @samp{yuv420}.
10458 @item rgb @emph{(deprecated)}
10459 If set to 1, force the filter to accept inputs in the RGB
10460 color space. Default value is 0. This option is deprecated, use
10461 @option{format} instead.
10464 If set to 1, force the filter to draw the last overlay frame over the
10465 main input until the end of the stream. A value of 0 disables this
10466 behavior. Default value is 1.
10469 The @option{x}, and @option{y} expressions can contain the following
10475 The main input width and height.
10479 The overlay input width and height.
10483 The computed values for @var{x} and @var{y}. They are evaluated for
10488 horizontal and vertical chroma subsample values of the output
10489 format. For example for the pixel format "yuv422p" @var{hsub} is 2 and
10493 the number of input frame, starting from 0
10496 the position in the file of the input frame, NAN if unknown
10499 The timestamp, expressed in seconds. It's NAN if the input timestamp is unknown.
10503 Note that the @var{n}, @var{pos}, @var{t} variables are available only
10504 when evaluation is done @emph{per frame}, and will evaluate to NAN
10505 when @option{eval} is set to @samp{init}.
10507 Be aware that frames are taken from each input video in timestamp
10508 order, hence, if their initial timestamps differ, it is a good idea
10509 to pass the two inputs through a @var{setpts=PTS-STARTPTS} filter to
10510 have them begin in the same zero timestamp, as the example for
10511 the @var{movie} filter does.
10513 You can chain together more overlays but you should test the
10514 efficiency of such approach.
10516 @subsection Commands
10518 This filter supports the following commands:
10522 Modify the x and y of the overlay input.
10523 The command accepts the same syntax of the corresponding option.
10525 If the specified expression is not valid, it is kept at its current
10529 @subsection Examples
10533 Draw the overlay at 10 pixels from the bottom right corner of the main
10536 overlay=main_w-overlay_w-10:main_h-overlay_h-10
10539 Using named options the example above becomes:
10541 overlay=x=main_w-overlay_w-10:y=main_h-overlay_h-10
10545 Insert a transparent PNG logo in the bottom left corner of the input,
10546 using the @command{ffmpeg} tool with the @code{-filter_complex} option:
10548 ffmpeg -i input -i logo -filter_complex 'overlay=10:main_h-overlay_h-10' output
10552 Insert 2 different transparent PNG logos (second logo on bottom
10553 right corner) using the @command{ffmpeg} tool:
10555 ffmpeg -i input -i logo1 -i logo2 -filter_complex 'overlay=x=10:y=H-h-10,overlay=x=W-w-10:y=H-h-10' output
10559 Add a transparent color layer on top of the main video; @code{WxH}
10560 must specify the size of the main input to the overlay filter:
10562 color=color=red@@.3:size=WxH [over]; [in][over] overlay [out]
10566 Play an original video and a filtered version (here with the deshake
10567 filter) side by side using the @command{ffplay} tool:
10569 ffplay input.avi -vf 'split[a][b]; [a]pad=iw*2:ih[src]; [b]deshake[filt]; [src][filt]overlay=w'
10572 The above command is the same as:
10574 ffplay input.avi -vf 'split[b], pad=iw*2[src], [b]deshake, [src]overlay=w'
10578 Make a sliding overlay appearing from the left to the right top part of the
10579 screen starting since time 2:
10581 overlay=x='if(gte(t,2), -w+(t-2)*20, NAN)':y=0
10585 Compose output by putting two input videos side to side:
10587 ffmpeg -i left.avi -i right.avi -filter_complex "
10588 nullsrc=size=200x100 [background];
10589 [0:v] setpts=PTS-STARTPTS, scale=100x100 [left];
10590 [1:v] setpts=PTS-STARTPTS, scale=100x100 [right];
10591 [background][left] overlay=shortest=1 [background+left];
10592 [background+left][right] overlay=shortest=1:x=100 [left+right]
10597 Mask 10-20 seconds of a video by applying the delogo filter to a section
10599 ffmpeg -i test.avi -codec:v:0 wmv2 -ar 11025 -b:v 9000k
10600 -vf '[in]split[split_main][split_delogo];[split_delogo]trim=start=360:end=371,delogo=0:0:640:480[delogoed];[split_main][delogoed]overlay=eof_action=pass[out]'
10605 Chain several overlays in cascade:
10607 nullsrc=s=200x200 [bg];
10608 testsrc=s=100x100, split=4 [in0][in1][in2][in3];
10609 [in0] lutrgb=r=0, [bg] overlay=0:0 [mid0];
10610 [in1] lutrgb=g=0, [mid0] overlay=100:0 [mid1];
10611 [in2] lutrgb=b=0, [mid1] overlay=0:100 [mid2];
10612 [in3] null, [mid2] overlay=100:100 [out0]
10619 Apply Overcomplete Wavelet denoiser.
10621 The filter accepts the following options:
10627 Larger depth values will denoise lower frequency components more, but
10628 slow down filtering.
10630 Must be an int in the range 8-16, default is @code{8}.
10632 @item luma_strength, ls
10635 Must be a double value in the range 0-1000, default is @code{1.0}.
10637 @item chroma_strength, cs
10638 Set chroma strength.
10640 Must be a double value in the range 0-1000, default is @code{1.0}.
10646 Add paddings to the input image, and place the original input at the
10647 provided @var{x}, @var{y} coordinates.
10649 It accepts the following parameters:
10654 Specify an expression for the size of the output image with the
10655 paddings added. If the value for @var{width} or @var{height} is 0, the
10656 corresponding input size is used for the output.
10658 The @var{width} expression can reference the value set by the
10659 @var{height} expression, and vice versa.
10661 The default value of @var{width} and @var{height} is 0.
10665 Specify the offsets to place the input image at within the padded area,
10666 with respect to the top/left border of the output image.
10668 The @var{x} expression can reference the value set by the @var{y}
10669 expression, and vice versa.
10671 The default value of @var{x} and @var{y} is 0.
10673 If @var{x} or @var{y} evaluate to a negative number, they'll be changed
10674 so the input image is centered on the padded area.
10677 Specify the color of the padded area. For the syntax of this option,
10678 check the "Color" section in the ffmpeg-utils manual.
10680 The default value of @var{color} is "black".
10683 Specify when to evaluate @var{width}, @var{height}, @var{x} and @var{y} expression.
10685 It accepts the following values:
10689 Only evaluate expressions once during the filter initialization or when
10690 a command is processed.
10693 Evaluate expressions for each incoming frame.
10697 Default value is @samp{init}.
10700 Pad to aspect instead to a resolution.
10704 The value for the @var{width}, @var{height}, @var{x}, and @var{y}
10705 options are expressions containing the following constants:
10710 The input video width and height.
10714 These are the same as @var{in_w} and @var{in_h}.
10718 The output width and height (the size of the padded area), as
10719 specified by the @var{width} and @var{height} expressions.
10723 These are the same as @var{out_w} and @var{out_h}.
10727 The x and y offsets as specified by the @var{x} and @var{y}
10728 expressions, or NAN if not yet specified.
10731 same as @var{iw} / @var{ih}
10734 input sample aspect ratio
10737 input display aspect ratio, it is the same as (@var{iw} / @var{ih}) * @var{sar}
10741 The horizontal and vertical chroma subsample values. For example for the
10742 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
10745 @subsection Examples
10749 Add paddings with the color "violet" to the input video. The output video
10750 size is 640x480, and the top-left corner of the input video is placed at
10753 pad=640:480:0:40:violet
10756 The example above is equivalent to the following command:
10758 pad=width=640:height=480:x=0:y=40:color=violet
10762 Pad the input to get an output with dimensions increased by 3/2,
10763 and put the input video at the center of the padded area:
10765 pad="3/2*iw:3/2*ih:(ow-iw)/2:(oh-ih)/2"
10769 Pad the input to get a squared output with size equal to the maximum
10770 value between the input width and height, and put the input video at
10771 the center of the padded area:
10773 pad="max(iw\,ih):ow:(ow-iw)/2:(oh-ih)/2"
10777 Pad the input to get a final w/h ratio of 16:9:
10779 pad="ih*16/9:ih:(ow-iw)/2:(oh-ih)/2"
10783 In case of anamorphic video, in order to set the output display aspect
10784 correctly, it is necessary to use @var{sar} in the expression,
10785 according to the relation:
10787 (ih * X / ih) * sar = output_dar
10788 X = output_dar / sar
10791 Thus the previous example needs to be modified to:
10793 pad="ih*16/9/sar:ih:(ow-iw)/2:(oh-ih)/2"
10797 Double the output size and put the input video in the bottom-right
10798 corner of the output padded area:
10800 pad="2*iw:2*ih:ow-iw:oh-ih"
10804 @anchor{palettegen}
10805 @section palettegen
10807 Generate one palette for a whole video stream.
10809 It accepts the following options:
10813 Set the maximum number of colors to quantize in the palette.
10814 Note: the palette will still contain 256 colors; the unused palette entries
10817 @item reserve_transparent
10818 Create a palette of 255 colors maximum and reserve the last one for
10819 transparency. Reserving the transparency color is useful for GIF optimization.
10820 If not set, the maximum of colors in the palette will be 256. You probably want
10821 to disable this option for a standalone image.
10825 Set statistics mode.
10827 It accepts the following values:
10830 Compute full frame histograms.
10832 Compute histograms only for the part that differs from previous frame. This
10833 might be relevant to give more importance to the moving part of your input if
10834 the background is static.
10836 Compute new histogram for each frame.
10839 Default value is @var{full}.
10842 The filter also exports the frame metadata @code{lavfi.color_quant_ratio}
10843 (@code{nb_color_in / nb_color_out}) which you can use to evaluate the degree of
10844 color quantization of the palette. This information is also visible at
10845 @var{info} logging level.
10847 @subsection Examples
10851 Generate a representative palette of a given video using @command{ffmpeg}:
10853 ffmpeg -i input.mkv -vf palettegen palette.png
10857 @section paletteuse
10859 Use a palette to downsample an input video stream.
10861 The filter takes two inputs: one video stream and a palette. The palette must
10862 be a 256 pixels image.
10864 It accepts the following options:
10868 Select dithering mode. Available algorithms are:
10871 Ordered 8x8 bayer dithering (deterministic)
10873 Dithering as defined by Paul Heckbert in 1982 (simple error diffusion).
10874 Note: this dithering is sometimes considered "wrong" and is included as a
10876 @item floyd_steinberg
10877 Floyd and Steingberg dithering (error diffusion)
10879 Frankie Sierra dithering v2 (error diffusion)
10881 Frankie Sierra dithering v2 "Lite" (error diffusion)
10884 Default is @var{sierra2_4a}.
10887 When @var{bayer} dithering is selected, this option defines the scale of the
10888 pattern (how much the crosshatch pattern is visible). A low value means more
10889 visible pattern for less banding, and higher value means less visible pattern
10890 at the cost of more banding.
10892 The option must be an integer value in the range [0,5]. Default is @var{2}.
10895 If set, define the zone to process
10899 Only the changing rectangle will be reprocessed. This is similar to GIF
10900 cropping/offsetting compression mechanism. This option can be useful for speed
10901 if only a part of the image is changing, and has use cases such as limiting the
10902 scope of the error diffusal @option{dither} to the rectangle that bounds the
10903 moving scene (it leads to more deterministic output if the scene doesn't change
10904 much, and as a result less moving noise and better GIF compression).
10907 Default is @var{none}.
10910 Take new palette for each output frame.
10913 @subsection Examples
10917 Use a palette (generated for example with @ref{palettegen}) to encode a GIF
10918 using @command{ffmpeg}:
10920 ffmpeg -i input.mkv -i palette.png -lavfi paletteuse output.gif
10924 @section perspective
10926 Correct perspective of video not recorded perpendicular to the screen.
10928 A description of the accepted parameters follows.
10939 Set coordinates expression for top left, top right, bottom left and bottom right corners.
10940 Default values are @code{0:0:W:0:0:H:W:H} with which perspective will remain unchanged.
10941 If the @code{sense} option is set to @code{source}, then the specified points will be sent
10942 to the corners of the destination. If the @code{sense} option is set to @code{destination},
10943 then the corners of the source will be sent to the specified coordinates.
10945 The expressions can use the following variables:
10950 the width and height of video frame.
10954 Output frame count.
10957 @item interpolation
10958 Set interpolation for perspective correction.
10960 It accepts the following values:
10966 Default value is @samp{linear}.
10969 Set interpretation of coordinate options.
10971 It accepts the following values:
10975 Send point in the source specified by the given coordinates to
10976 the corners of the destination.
10978 @item 1, destination
10980 Send the corners of the source to the point in the destination specified
10981 by the given coordinates.
10983 Default value is @samp{source}.
10987 Set when the expressions for coordinates @option{x0,y0,...x3,y3} are evaluated.
10989 It accepts the following values:
10992 only evaluate expressions once during the filter initialization or
10993 when a command is processed
10996 evaluate expressions for each incoming frame
10999 Default value is @samp{init}.
11004 Delay interlaced video by one field time so that the field order changes.
11006 The intended use is to fix PAL movies that have been captured with the
11007 opposite field order to the film-to-video transfer.
11009 A description of the accepted parameters follows.
11015 It accepts the following values:
11018 Capture field order top-first, transfer bottom-first.
11019 Filter will delay the bottom field.
11022 Capture field order bottom-first, transfer top-first.
11023 Filter will delay the top field.
11026 Capture and transfer with the same field order. This mode only exists
11027 for the documentation of the other options to refer to, but if you
11028 actually select it, the filter will faithfully do nothing.
11031 Capture field order determined automatically by field flags, transfer
11033 Filter selects among @samp{t} and @samp{b} modes on a frame by frame
11034 basis using field flags. If no field information is available,
11035 then this works just like @samp{u}.
11038 Capture unknown or varying, transfer opposite.
11039 Filter selects among @samp{t} and @samp{b} on a frame by frame basis by
11040 analyzing the images and selecting the alternative that produces best
11041 match between the fields.
11044 Capture top-first, transfer unknown or varying.
11045 Filter selects among @samp{t} and @samp{p} using image analysis.
11048 Capture bottom-first, transfer unknown or varying.
11049 Filter selects among @samp{b} and @samp{p} using image analysis.
11052 Capture determined by field flags, transfer unknown or varying.
11053 Filter selects among @samp{t}, @samp{b} and @samp{p} using field flags and
11054 image analysis. If no field information is available, then this works just
11055 like @samp{U}. This is the default mode.
11058 Both capture and transfer unknown or varying.
11059 Filter selects among @samp{t}, @samp{b} and @samp{p} using image analysis only.
11063 @section pixdesctest
11065 Pixel format descriptor test filter, mainly useful for internal
11066 testing. The output video should be equal to the input video.
11070 format=monow, pixdesctest
11073 can be used to test the monowhite pixel format descriptor definition.
11077 Display sample values of color channels. Mainly useful for checking color and levels.
11079 The filters accept the following options:
11083 Set scope X position, offset on X axis.
11086 Set scope Y position, offset on Y axis.
11095 Set window opacity. This window also holds statistics about pixel area.
11100 Enable the specified chain of postprocessing subfilters using libpostproc. This
11101 library should be automatically selected with a GPL build (@code{--enable-gpl}).
11102 Subfilters must be separated by '/' and can be disabled by prepending a '-'.
11103 Each subfilter and some options have a short and a long name that can be used
11104 interchangeably, i.e. dr/dering are the same.
11106 The filters accept the following options:
11110 Set postprocessing subfilters string.
11113 All subfilters share common options to determine their scope:
11117 Honor the quality commands for this subfilter.
11120 Do chrominance filtering, too (default).
11123 Do luminance filtering only (no chrominance).
11126 Do chrominance filtering only (no luminance).
11129 These options can be appended after the subfilter name, separated by a '|'.
11131 Available subfilters are:
11134 @item hb/hdeblock[|difference[|flatness]]
11135 Horizontal deblocking filter
11138 Difference factor where higher values mean more deblocking (default: @code{32}).
11140 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11143 @item vb/vdeblock[|difference[|flatness]]
11144 Vertical deblocking filter
11147 Difference factor where higher values mean more deblocking (default: @code{32}).
11149 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11152 @item ha/hadeblock[|difference[|flatness]]
11153 Accurate horizontal deblocking filter
11156 Difference factor where higher values mean more deblocking (default: @code{32}).
11158 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11161 @item va/vadeblock[|difference[|flatness]]
11162 Accurate vertical deblocking filter
11165 Difference factor where higher values mean more deblocking (default: @code{32}).
11167 Flatness threshold where lower values mean more deblocking (default: @code{39}).
11171 The horizontal and vertical deblocking filters share the difference and
11172 flatness values so you cannot set different horizontal and vertical
11176 @item h1/x1hdeblock
11177 Experimental horizontal deblocking filter
11179 @item v1/x1vdeblock
11180 Experimental vertical deblocking filter
11185 @item tn/tmpnoise[|threshold1[|threshold2[|threshold3]]], temporal noise reducer
11188 larger -> stronger filtering
11190 larger -> stronger filtering
11192 larger -> stronger filtering
11195 @item al/autolevels[:f/fullyrange], automatic brightness / contrast correction
11198 Stretch luminance to @code{0-255}.
11201 @item lb/linblenddeint
11202 Linear blend deinterlacing filter that deinterlaces the given block by
11203 filtering all lines with a @code{(1 2 1)} filter.
11205 @item li/linipoldeint
11206 Linear interpolating deinterlacing filter that deinterlaces the given block by
11207 linearly interpolating every second line.
11209 @item ci/cubicipoldeint
11210 Cubic interpolating deinterlacing filter deinterlaces the given block by
11211 cubically interpolating every second line.
11213 @item md/mediandeint
11214 Median deinterlacing filter that deinterlaces the given block by applying a
11215 median filter to every second line.
11217 @item fd/ffmpegdeint
11218 FFmpeg deinterlacing filter that deinterlaces the given block by filtering every
11219 second line with a @code{(-1 4 2 4 -1)} filter.
11222 Vertically applied FIR lowpass deinterlacing filter that deinterlaces the given
11223 block by filtering all lines with a @code{(-1 2 6 2 -1)} filter.
11225 @item fq/forceQuant[|quantizer]
11226 Overrides the quantizer table from the input with the constant quantizer you
11234 Default pp filter combination (@code{hb|a,vb|a,dr|a})
11237 Fast pp filter combination (@code{h1|a,v1|a,dr|a})
11240 High quality pp filter combination (@code{ha|a|128|7,va|a,dr|a})
11243 @subsection Examples
11247 Apply horizontal and vertical deblocking, deringing and automatic
11248 brightness/contrast:
11254 Apply default filters without brightness/contrast correction:
11260 Apply default filters and temporal denoiser:
11262 pp=default/tmpnoise|1|2|3
11266 Apply deblocking on luminance only, and switch vertical deblocking on or off
11267 automatically depending on available CPU time:
11274 Apply Postprocessing filter 7. It is variant of the @ref{spp} filter,
11275 similar to spp = 6 with 7 point DCT, where only the center sample is
11278 The filter accepts the following options:
11282 Force a constant quantization parameter. It accepts an integer in range
11283 0 to 63. If not set, the filter will use the QP from the video stream
11287 Set thresholding mode. Available modes are:
11291 Set hard thresholding.
11293 Set soft thresholding (better de-ringing effect, but likely blurrier).
11295 Set medium thresholding (good results, default).
11299 @section premultiply
11300 Apply alpha premultiply effect to input video stream using first plane
11301 of second stream as alpha.
11303 Both streams must have same dimensions and same pixel format.
11305 The filter accepts the following option:
11309 Set which planes will be processed, unprocessed planes will be copied.
11310 By default value 0xf, all planes will be processed.
11314 Apply prewitt operator to input video stream.
11316 The filter accepts the following option:
11320 Set which planes will be processed, unprocessed planes will be copied.
11321 By default value 0xf, all planes will be processed.
11324 Set value which will be multiplied with filtered result.
11327 Set value which will be added to filtered result.
11332 Obtain the average, maximum and minimum PSNR (Peak Signal to Noise
11333 Ratio) between two input videos.
11335 This filter takes in input two input videos, the first input is
11336 considered the "main" source and is passed unchanged to the
11337 output. The second input is used as a "reference" video for computing
11340 Both video inputs must have the same resolution and pixel format for
11341 this filter to work correctly. Also it assumes that both inputs
11342 have the same number of frames, which are compared one by one.
11344 The obtained average PSNR is printed through the logging system.
11346 The filter stores the accumulated MSE (mean squared error) of each
11347 frame, and at the end of the processing it is averaged across all frames
11348 equally, and the following formula is applied to obtain the PSNR:
11351 PSNR = 10*log10(MAX^2/MSE)
11354 Where MAX is the average of the maximum values of each component of the
11357 The description of the accepted parameters follows.
11360 @item stats_file, f
11361 If specified the filter will use the named file to save the PSNR of
11362 each individual frame. When filename equals "-" the data is sent to
11365 @item stats_version
11366 Specifies which version of the stats file format to use. Details of
11367 each format are written below.
11368 Default value is 1.
11370 @item stats_add_max
11371 Determines whether the max value is output to the stats log.
11372 Default value is 0.
11373 Requires stats_version >= 2. If this is set and stats_version < 2,
11374 the filter will return an error.
11377 The file printed if @var{stats_file} is selected, contains a sequence of
11378 key/value pairs of the form @var{key}:@var{value} for each compared
11381 If a @var{stats_version} greater than 1 is specified, a header line precedes
11382 the list of per-frame-pair stats, with key value pairs following the frame
11383 format with the following parameters:
11386 @item psnr_log_version
11387 The version of the log file format. Will match @var{stats_version}.
11390 A comma separated list of the per-frame-pair parameters included in
11394 A description of each shown per-frame-pair parameter follows:
11398 sequential number of the input frame, starting from 1
11401 Mean Square Error pixel-by-pixel average difference of the compared
11402 frames, averaged over all the image components.
11404 @item mse_y, mse_u, mse_v, mse_r, mse_g, mse_g, mse_a
11405 Mean Square Error pixel-by-pixel average difference of the compared
11406 frames for the component specified by the suffix.
11408 @item psnr_y, psnr_u, psnr_v, psnr_r, psnr_g, psnr_b, psnr_a
11409 Peak Signal to Noise ratio of the compared frames for the component
11410 specified by the suffix.
11412 @item max_avg, max_y, max_u, max_v
11413 Maximum allowed value for each channel, and average over all
11419 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
11420 [main][ref] psnr="stats_file=stats.log" [out]
11423 On this example the input file being processed is compared with the
11424 reference file @file{ref_movie.mpg}. The PSNR of each individual frame
11425 is stored in @file{stats.log}.
11430 Pulldown reversal (inverse telecine) filter, capable of handling mixed
11431 hard-telecine, 24000/1001 fps progressive, and 30000/1001 fps progressive
11434 The pullup filter is designed to take advantage of future context in making
11435 its decisions. This filter is stateless in the sense that it does not lock
11436 onto a pattern to follow, but it instead looks forward to the following
11437 fields in order to identify matches and rebuild progressive frames.
11439 To produce content with an even framerate, insert the fps filter after
11440 pullup, use @code{fps=24000/1001} if the input frame rate is 29.97fps,
11441 @code{fps=24} for 30fps and the (rare) telecined 25fps input.
11443 The filter accepts the following options:
11450 These options set the amount of "junk" to ignore at the left, right, top, and
11451 bottom of the image, respectively. Left and right are in units of 8 pixels,
11452 while top and bottom are in units of 2 lines.
11453 The default is 8 pixels on each side.
11456 Set the strict breaks. Setting this option to 1 will reduce the chances of
11457 filter generating an occasional mismatched frame, but it may also cause an
11458 excessive number of frames to be dropped during high motion sequences.
11459 Conversely, setting it to -1 will make filter match fields more easily.
11460 This may help processing of video where there is slight blurring between
11461 the fields, but may also cause there to be interlaced frames in the output.
11462 Default value is @code{0}.
11465 Set the metric plane to use. It accepts the following values:
11471 Use chroma blue plane.
11474 Use chroma red plane.
11477 This option may be set to use chroma plane instead of the default luma plane
11478 for doing filter's computations. This may improve accuracy on very clean
11479 source material, but more likely will decrease accuracy, especially if there
11480 is chroma noise (rainbow effect) or any grayscale video.
11481 The main purpose of setting @option{mp} to a chroma plane is to reduce CPU
11482 load and make pullup usable in realtime on slow machines.
11485 For best results (without duplicated frames in the output file) it is
11486 necessary to change the output frame rate. For example, to inverse
11487 telecine NTSC input:
11489 ffmpeg -i input -vf pullup -r 24000/1001 ...
11494 Change video quantization parameters (QP).
11496 The filter accepts the following option:
11500 Set expression for quantization parameter.
11503 The expression is evaluated through the eval API and can contain, among others,
11504 the following constants:
11508 1 if index is not 129, 0 otherwise.
11511 Sequentional index starting from -129 to 128.
11514 @subsection Examples
11518 Some equation like:
11526 Flush video frames from internal cache of frames into a random order.
11527 No frame is discarded.
11528 Inspired by @ref{frei0r} nervous filter.
11532 Set size in number of frames of internal cache, in range from @code{2} to
11533 @code{512}. Default is @code{30}.
11536 Set seed for random number generator, must be an integer included between
11537 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
11538 less than @code{0}, the filter will try to use a good random seed on a
11542 @section readeia608
11544 Read closed captioning (EIA-608) information from the top lines of a video frame.
11546 This filter adds frame metadata for @code{lavfi.readeia608.X.cc} and
11547 @code{lavfi.readeia608.X.line}, where @code{X} is the number of the identified line
11548 with EIA-608 data (starting from 0). A description of each metadata value follows:
11551 @item lavfi.readeia608.X.cc
11552 The two bytes stored as EIA-608 data (printed in hexadecimal).
11554 @item lavfi.readeia608.X.line
11555 The number of the line on which the EIA-608 data was identified and read.
11558 This filter accepts the following options:
11562 Set the line to start scanning for EIA-608 data. Default is @code{0}.
11565 Set the line to end scanning for EIA-608 data. Default is @code{29}.
11568 Set minimal acceptable amplitude change for sync codes detection.
11569 Default is @code{0.2}. Allowed range is @code{[0.001 - 1]}.
11572 Set the ratio of width reserved for sync code detection.
11573 Default is @code{0.27}. Allowed range is @code{[0.01 - 0.7]}.
11576 Set the max peaks height difference for sync code detection.
11577 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
11580 Set max peaks period difference for sync code detection.
11581 Default is @code{0.1}. Allowed range is @code{[0.0 - 0.5]}.
11584 Set the first two max start code bits differences.
11585 Default is @code{0.02}. Allowed range is @code{[0.0 - 0.5]}.
11588 Set the minimum ratio of bits height compared to 3rd start code bit.
11589 Default is @code{0.75}. Allowed range is @code{[0.01 - 1]}.
11592 Set the white color threshold. Default is @code{0.35}. Allowed range is @code{[0.1 - 1]}.
11595 Set the black color threshold. Default is @code{0.15}. Allowed range is @code{[0.0 - 0.5]}.
11598 Enable checking the parity bit. In the event of a parity error, the filter will output
11599 @code{0x00} for that character. Default is false.
11602 @subsection Examples
11606 Output a csv with presentation time and the first two lines of identified EIA-608 captioning data.
11608 ffprobe -f lavfi -i movie=captioned_video.mov,readeia608 -show_entries frame=pkt_pts_time:frame_tags=lavfi.readeia608.0.cc,lavfi.readeia608.1.cc -of csv
11614 Read vertical interval timecode (VITC) information from the top lines of a
11617 The filter adds frame metadata key @code{lavfi.readvitc.tc_str} with the
11618 timecode value, if a valid timecode has been detected. Further metadata key
11619 @code{lavfi.readvitc.found} is set to 0/1 depending on whether
11620 timecode data has been found or not.
11622 This filter accepts the following options:
11626 Set the maximum number of lines to scan for VITC data. If the value is set to
11627 @code{-1} the full video frame is scanned. Default is @code{45}.
11630 Set the luma threshold for black. Accepts float numbers in the range [0.0,1.0],
11631 default value is @code{0.2}. The value must be equal or less than @code{thr_w}.
11634 Set the luma threshold for white. Accepts float numbers in the range [0.0,1.0],
11635 default value is @code{0.6}. The value must be equal or greater than @code{thr_b}.
11638 @subsection Examples
11642 Detect and draw VITC data onto the video frame; if no valid VITC is detected,
11643 draw @code{--:--:--:--} as a placeholder:
11645 ffmpeg -i input.avi -filter:v 'readvitc,drawtext=fontfile=FreeMono.ttf:text=%@{metadata\\:lavfi.readvitc.tc_str\\:--\\\\\\:--\\\\\\:--\\\\\\:--@}:x=(w-tw)/2:y=400-ascent'
11651 Remap pixels using 2nd: Xmap and 3rd: Ymap input video stream.
11653 Destination pixel at position (X, Y) will be picked from source (x, y) position
11654 where x = Xmap(X, Y) and y = Ymap(X, Y). If mapping values are out of range, zero
11655 value for pixel will be used for destination pixel.
11657 Xmap and Ymap input video streams must be of same dimensions. Output video stream
11658 will have Xmap/Ymap video stream dimensions.
11659 Xmap and Ymap input video streams are 16bit depth, single channel.
11661 @section removegrain
11663 The removegrain filter is a spatial denoiser for progressive video.
11667 Set mode for the first plane.
11670 Set mode for the second plane.
11673 Set mode for the third plane.
11676 Set mode for the fourth plane.
11679 Range of mode is from 0 to 24. Description of each mode follows:
11683 Leave input plane unchanged. Default.
11686 Clips the pixel with the minimum and maximum of the 8 neighbour pixels.
11689 Clips the pixel with the second minimum and maximum of the 8 neighbour pixels.
11692 Clips the pixel with the third minimum and maximum of the 8 neighbour pixels.
11695 Clips the pixel with the fourth minimum and maximum of the 8 neighbour pixels.
11696 This is equivalent to a median filter.
11699 Line-sensitive clipping giving the minimal change.
11702 Line-sensitive clipping, intermediate.
11705 Line-sensitive clipping, intermediate.
11708 Line-sensitive clipping, intermediate.
11711 Line-sensitive clipping on a line where the neighbours pixels are the closest.
11714 Replaces the target pixel with the closest neighbour.
11717 [1 2 1] horizontal and vertical kernel blur.
11723 Bob mode, interpolates top field from the line where the neighbours
11724 pixels are the closest.
11727 Bob mode, interpolates bottom field from the line where the neighbours
11728 pixels are the closest.
11731 Bob mode, interpolates top field. Same as 13 but with a more complicated
11732 interpolation formula.
11735 Bob mode, interpolates bottom field. Same as 14 but with a more complicated
11736 interpolation formula.
11739 Clips the pixel with the minimum and maximum of respectively the maximum and
11740 minimum of each pair of opposite neighbour pixels.
11743 Line-sensitive clipping using opposite neighbours whose greatest distance from
11744 the current pixel is minimal.
11747 Replaces the pixel with the average of its 8 neighbours.
11750 Averages the 9 pixels ([1 1 1] horizontal and vertical blur).
11753 Clips pixels using the averages of opposite neighbour.
11756 Same as mode 21 but simpler and faster.
11759 Small edge and halo removal, but reputed useless.
11765 @section removelogo
11767 Suppress a TV station logo, using an image file to determine which
11768 pixels comprise the logo. It works by filling in the pixels that
11769 comprise the logo with neighboring pixels.
11771 The filter accepts the following options:
11775 Set the filter bitmap file, which can be any image format supported by
11776 libavformat. The width and height of the image file must match those of the
11777 video stream being processed.
11780 Pixels in the provided bitmap image with a value of zero are not
11781 considered part of the logo, non-zero pixels are considered part of
11782 the logo. If you use white (255) for the logo and black (0) for the
11783 rest, you will be safe. For making the filter bitmap, it is
11784 recommended to take a screen capture of a black frame with the logo
11785 visible, and then using a threshold filter followed by the erode
11786 filter once or twice.
11788 If needed, little splotches can be fixed manually. Remember that if
11789 logo pixels are not covered, the filter quality will be much
11790 reduced. Marking too many pixels as part of the logo does not hurt as
11791 much, but it will increase the amount of blurring needed to cover over
11792 the image and will destroy more information than necessary, and extra
11793 pixels will slow things down on a large logo.
11795 @section repeatfields
11797 This filter uses the repeat_field flag from the Video ES headers and hard repeats
11798 fields based on its value.
11802 Reverse a video clip.
11804 Warning: This filter requires memory to buffer the entire clip, so trimming
11807 @subsection Examples
11811 Take the first 5 seconds of a clip, and reverse it.
11819 Rotate video by an arbitrary angle expressed in radians.
11821 The filter accepts the following options:
11823 A description of the optional parameters follows.
11826 Set an expression for the angle by which to rotate the input video
11827 clockwise, expressed as a number of radians. A negative value will
11828 result in a counter-clockwise rotation. By default it is set to "0".
11830 This expression is evaluated for each frame.
11833 Set the output width expression, default value is "iw".
11834 This expression is evaluated just once during configuration.
11837 Set the output height expression, default value is "ih".
11838 This expression is evaluated just once during configuration.
11841 Enable bilinear interpolation if set to 1, a value of 0 disables
11842 it. Default value is 1.
11845 Set the color used to fill the output area not covered by the rotated
11846 image. For the general syntax of this option, check the "Color" section in the
11847 ffmpeg-utils manual. If the special value "none" is selected then no
11848 background is printed (useful for example if the background is never shown).
11850 Default value is "black".
11853 The expressions for the angle and the output size can contain the
11854 following constants and functions:
11858 sequential number of the input frame, starting from 0. It is always NAN
11859 before the first frame is filtered.
11862 time in seconds of the input frame, it is set to 0 when the filter is
11863 configured. It is always NAN before the first frame is filtered.
11867 horizontal and vertical chroma subsample values. For example for the
11868 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
11872 the input video width and height
11876 the output width and height, that is the size of the padded area as
11877 specified by the @var{width} and @var{height} expressions
11881 the minimal width/height required for completely containing the input
11882 video rotated by @var{a} radians.
11884 These are only available when computing the @option{out_w} and
11885 @option{out_h} expressions.
11888 @subsection Examples
11892 Rotate the input by PI/6 radians clockwise:
11898 Rotate the input by PI/6 radians counter-clockwise:
11904 Rotate the input by 45 degrees clockwise:
11910 Apply a constant rotation with period T, starting from an angle of PI/3:
11912 rotate=PI/3+2*PI*t/T
11916 Make the input video rotation oscillating with a period of T
11917 seconds and an amplitude of A radians:
11919 rotate=A*sin(2*PI/T*t)
11923 Rotate the video, output size is chosen so that the whole rotating
11924 input video is always completely contained in the output:
11926 rotate='2*PI*t:ow=hypot(iw,ih):oh=ow'
11930 Rotate the video, reduce the output size so that no background is ever
11933 rotate=2*PI*t:ow='min(iw,ih)/sqrt(2)':oh=ow:c=none
11937 @subsection Commands
11939 The filter supports the following commands:
11943 Set the angle expression.
11944 The command accepts the same syntax of the corresponding option.
11946 If the specified expression is not valid, it is kept at its current
11952 Apply Shape Adaptive Blur.
11954 The filter accepts the following options:
11957 @item luma_radius, lr
11958 Set luma blur filter strength, must be a value in range 0.1-4.0, default
11959 value is 1.0. A greater value will result in a more blurred image, and
11960 in slower processing.
11962 @item luma_pre_filter_radius, lpfr
11963 Set luma pre-filter radius, must be a value in the 0.1-2.0 range, default
11966 @item luma_strength, ls
11967 Set luma maximum difference between pixels to still be considered, must
11968 be a value in the 0.1-100.0 range, default value is 1.0.
11970 @item chroma_radius, cr
11971 Set chroma blur filter strength, must be a value in range -0.9-4.0. A
11972 greater value will result in a more blurred image, and in slower
11975 @item chroma_pre_filter_radius, cpfr
11976 Set chroma pre-filter radius, must be a value in the -0.9-2.0 range.
11978 @item chroma_strength, cs
11979 Set chroma maximum difference between pixels to still be considered,
11980 must be a value in the -0.9-100.0 range.
11983 Each chroma option value, if not explicitly specified, is set to the
11984 corresponding luma option value.
11989 Scale (resize) the input video, using the libswscale library.
11991 The scale filter forces the output display aspect ratio to be the same
11992 of the input, by changing the output sample aspect ratio.
11994 If the input image format is different from the format requested by
11995 the next filter, the scale filter will convert the input to the
11998 @subsection Options
11999 The filter accepts the following options, or any of the options
12000 supported by the libswscale scaler.
12002 See @ref{scaler_options,,the ffmpeg-scaler manual,ffmpeg-scaler} for
12003 the complete list of scaler options.
12008 Set the output video dimension expression. Default value is the input
12011 If the value is 0, the input width is used for the output.
12013 If one of the values is -1, the scale filter will use a value that
12014 maintains the aspect ratio of the input image, calculated from the
12015 other specified dimension. If both of them are -1, the input size is
12018 If one of the values is -n with n > 1, the scale filter will also use a value
12019 that maintains the aspect ratio of the input image, calculated from the other
12020 specified dimension. After that it will, however, make sure that the calculated
12021 dimension is divisible by n and adjust the value if necessary.
12023 See below for the list of accepted constants for use in the dimension
12027 Specify when to evaluate @var{width} and @var{height} expression. It accepts the following values:
12031 Only evaluate expressions once during the filter initialization or when a command is processed.
12034 Evaluate expressions for each incoming frame.
12038 Default value is @samp{init}.
12042 Set the interlacing mode. It accepts the following values:
12046 Force interlaced aware scaling.
12049 Do not apply interlaced scaling.
12052 Select interlaced aware scaling depending on whether the source frames
12053 are flagged as interlaced or not.
12056 Default value is @samp{0}.
12059 Set libswscale scaling flags. See
12060 @ref{sws_flags,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12061 complete list of values. If not explicitly specified the filter applies
12065 @item param0, param1
12066 Set libswscale input parameters for scaling algorithms that need them. See
12067 @ref{sws_params,,the ffmpeg-scaler manual,ffmpeg-scaler} for the
12068 complete documentation. If not explicitly specified the filter applies
12074 Set the video size. For the syntax of this option, check the
12075 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12077 @item in_color_matrix
12078 @item out_color_matrix
12079 Set in/output YCbCr color space type.
12081 This allows the autodetected value to be overridden as well as allows forcing
12082 a specific value used for the output and encoder.
12084 If not specified, the color space type depends on the pixel format.
12090 Choose automatically.
12093 Format conforming to International Telecommunication Union (ITU)
12094 Recommendation BT.709.
12097 Set color space conforming to the United States Federal Communications
12098 Commission (FCC) Code of Federal Regulations (CFR) Title 47 (2003) 73.682 (a).
12101 Set color space conforming to:
12105 ITU Radiocommunication Sector (ITU-R) Recommendation BT.601
12108 ITU-R Rec. BT.470-6 (1998) Systems B, B1, and G
12111 Society of Motion Picture and Television Engineers (SMPTE) ST 170:2004
12116 Set color space conforming to SMPTE ST 240:1999.
12121 Set in/output YCbCr sample range.
12123 This allows the autodetected value to be overridden as well as allows forcing
12124 a specific value used for the output and encoder. If not specified, the
12125 range depends on the pixel format. Possible values:
12129 Choose automatically.
12132 Set full range (0-255 in case of 8-bit luma).
12135 Set "MPEG" range (16-235 in case of 8-bit luma).
12138 @item force_original_aspect_ratio
12139 Enable decreasing or increasing output video width or height if necessary to
12140 keep the original aspect ratio. Possible values:
12144 Scale the video as specified and disable this feature.
12147 The output video dimensions will automatically be decreased if needed.
12150 The output video dimensions will automatically be increased if needed.
12154 One useful instance of this option is that when you know a specific device's
12155 maximum allowed resolution, you can use this to limit the output video to
12156 that, while retaining the aspect ratio. For example, device A allows
12157 1280x720 playback, and your video is 1920x800. Using this option (set it to
12158 decrease) and specifying 1280x720 to the command line makes the output
12161 Please note that this is a different thing than specifying -1 for @option{w}
12162 or @option{h}, you still need to specify the output resolution for this option
12167 The values of the @option{w} and @option{h} options are expressions
12168 containing the following constants:
12173 The input width and height
12177 These are the same as @var{in_w} and @var{in_h}.
12181 The output (scaled) width and height
12185 These are the same as @var{out_w} and @var{out_h}
12188 The same as @var{iw} / @var{ih}
12191 input sample aspect ratio
12194 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
12198 horizontal and vertical input chroma subsample values. For example for the
12199 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12203 horizontal and vertical output chroma subsample values. For example for the
12204 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12207 @subsection Examples
12211 Scale the input video to a size of 200x100
12216 This is equivalent to:
12227 Specify a size abbreviation for the output size:
12232 which can also be written as:
12238 Scale the input to 2x:
12240 scale=w=2*iw:h=2*ih
12244 The above is the same as:
12246 scale=2*in_w:2*in_h
12250 Scale the input to 2x with forced interlaced scaling:
12252 scale=2*iw:2*ih:interl=1
12256 Scale the input to half size:
12258 scale=w=iw/2:h=ih/2
12262 Increase the width, and set the height to the same size:
12268 Seek Greek harmony:
12275 Increase the height, and set the width to 3/2 of the height:
12277 scale=w=3/2*oh:h=3/5*ih
12281 Increase the size, making the size a multiple of the chroma
12284 scale="trunc(3/2*iw/hsub)*hsub:trunc(3/2*ih/vsub)*vsub"
12288 Increase the width to a maximum of 500 pixels,
12289 keeping the same aspect ratio as the input:
12291 scale=w='min(500\, iw*3/2):h=-1'
12295 @subsection Commands
12297 This filter supports the following commands:
12301 Set the output video dimension expression.
12302 The command accepts the same syntax of the corresponding option.
12304 If the specified expression is not valid, it is kept at its current
12310 Use the NVIDIA Performance Primitives (libnpp) to perform scaling and/or pixel
12311 format conversion on CUDA video frames. Setting the output width and height
12312 works in the same way as for the @var{scale} filter.
12314 The following additional options are accepted:
12317 The pixel format of the output CUDA frames. If set to the string "same" (the
12318 default), the input format will be kept. Note that automatic format negotiation
12319 and conversion is not yet supported for hardware frames
12322 The interpolation algorithm used for resizing. One of the following:
12329 @item cubic2p_bspline
12330 2-parameter cubic (B=1, C=0)
12332 @item cubic2p_catmullrom
12333 2-parameter cubic (B=0, C=1/2)
12335 @item cubic2p_b05c03
12336 2-parameter cubic (B=1/2, C=3/10)
12348 Scale (resize) the input video, based on a reference video.
12350 See the scale filter for available options, scale2ref supports the same but
12351 uses the reference video instead of the main input as basis.
12353 @subsection Examples
12357 Scale a subtitle stream to match the main video in size before overlaying
12359 'scale2ref[b][a];[a][b]overlay'
12363 @anchor{selectivecolor}
12364 @section selectivecolor
12366 Adjust cyan, magenta, yellow and black (CMYK) to certain ranges of colors (such
12367 as "reds", "yellows", "greens", "cyans", ...). The adjustment range is defined
12368 by the "purity" of the color (that is, how saturated it already is).
12370 This filter is similar to the Adobe Photoshop Selective Color tool.
12372 The filter accepts the following options:
12375 @item correction_method
12376 Select color correction method.
12378 Available values are:
12381 Specified adjustments are applied "as-is" (added/subtracted to original pixel
12384 Specified adjustments are relative to the original component value.
12386 Default is @code{absolute}.
12388 Adjustments for red pixels (pixels where the red component is the maximum)
12390 Adjustments for yellow pixels (pixels where the blue component is the minimum)
12392 Adjustments for green pixels (pixels where the green component is the maximum)
12394 Adjustments for cyan pixels (pixels where the red component is the minimum)
12396 Adjustments for blue pixels (pixels where the blue component is the maximum)
12398 Adjustments for magenta pixels (pixels where the green component is the minimum)
12400 Adjustments for white pixels (pixels where all components are greater than 128)
12402 Adjustments for all pixels except pure black and pure white
12404 Adjustments for black pixels (pixels where all components are lesser than 128)
12406 Specify a Photoshop selective color file (@code{.asv}) to import the settings from.
12409 All the adjustment settings (@option{reds}, @option{yellows}, ...) accept up to
12410 4 space separated floating point adjustment values in the [-1,1] range,
12411 respectively to adjust the amount of cyan, magenta, yellow and black for the
12412 pixels of its range.
12414 @subsection Examples
12418 Increase cyan by 50% and reduce yellow by 33% in every green areas, and
12419 increase magenta by 27% in blue areas:
12421 selectivecolor=greens=.5 0 -.33 0:blues=0 .27
12425 Use a Photoshop selective color preset:
12427 selectivecolor=psfile=MySelectiveColorPresets/Misty.asv
12431 @anchor{separatefields}
12432 @section separatefields
12434 The @code{separatefields} takes a frame-based video input and splits
12435 each frame into its components fields, producing a new half height clip
12436 with twice the frame rate and twice the frame count.
12438 This filter use field-dominance information in frame to decide which
12439 of each pair of fields to place first in the output.
12440 If it gets it wrong use @ref{setfield} filter before @code{separatefields} filter.
12442 @section setdar, setsar
12444 The @code{setdar} filter sets the Display Aspect Ratio for the filter
12447 This is done by changing the specified Sample (aka Pixel) Aspect
12448 Ratio, according to the following equation:
12450 @var{DAR} = @var{HORIZONTAL_RESOLUTION} / @var{VERTICAL_RESOLUTION} * @var{SAR}
12453 Keep in mind that the @code{setdar} filter does not modify the pixel
12454 dimensions of the video frame. Also, the display aspect ratio set by
12455 this filter may be changed by later filters in the filterchain,
12456 e.g. in case of scaling or if another "setdar" or a "setsar" filter is
12459 The @code{setsar} filter sets the Sample (aka Pixel) Aspect Ratio for
12460 the filter output video.
12462 Note that as a consequence of the application of this filter, the
12463 output display aspect ratio will change according to the equation
12466 Keep in mind that the sample aspect ratio set by the @code{setsar}
12467 filter may be changed by later filters in the filterchain, e.g. if
12468 another "setsar" or a "setdar" filter is applied.
12470 It accepts the following parameters:
12473 @item r, ratio, dar (@code{setdar} only), sar (@code{setsar} only)
12474 Set the aspect ratio used by the filter.
12476 The parameter can be a floating point number string, an expression, or
12477 a string of the form @var{num}:@var{den}, where @var{num} and
12478 @var{den} are the numerator and denominator of the aspect ratio. If
12479 the parameter is not specified, it is assumed the value "0".
12480 In case the form "@var{num}:@var{den}" is used, the @code{:} character
12484 Set the maximum integer value to use for expressing numerator and
12485 denominator when reducing the expressed aspect ratio to a rational.
12486 Default value is @code{100}.
12490 The parameter @var{sar} is an expression containing
12491 the following constants:
12495 These are approximated values for the mathematical constants e
12496 (Euler's number), pi (Greek pi), and phi (the golden ratio).
12499 The input width and height.
12502 These are the same as @var{w} / @var{h}.
12505 The input sample aspect ratio.
12508 The input display aspect ratio. It is the same as
12509 (@var{w} / @var{h}) * @var{sar}.
12512 Horizontal and vertical chroma subsample values. For example, for the
12513 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
12516 @subsection Examples
12521 To change the display aspect ratio to 16:9, specify one of the following:
12528 To change the sample aspect ratio to 10:11, specify:
12534 To set a display aspect ratio of 16:9, and specify a maximum integer value of
12535 1000 in the aspect ratio reduction, use the command:
12537 setdar=ratio=16/9:max=1000
12545 Force field for the output video frame.
12547 The @code{setfield} filter marks the interlace type field for the
12548 output frames. It does not change the input frame, but only sets the
12549 corresponding property, which affects how the frame is treated by
12550 following filters (e.g. @code{fieldorder} or @code{yadif}).
12552 The filter accepts the following options:
12557 Available values are:
12561 Keep the same field property.
12564 Mark the frame as bottom-field-first.
12567 Mark the frame as top-field-first.
12570 Mark the frame as progressive.
12576 Show a line containing various information for each input video frame.
12577 The input video is not modified.
12579 The shown line contains a sequence of key/value pairs of the form
12580 @var{key}:@var{value}.
12582 The following values are shown in the output:
12586 The (sequential) number of the input frame, starting from 0.
12589 The Presentation TimeStamp of the input frame, expressed as a number of
12590 time base units. The time base unit depends on the filter input pad.
12593 The Presentation TimeStamp of the input frame, expressed as a number of
12597 The position of the frame in the input stream, or -1 if this information is
12598 unavailable and/or meaningless (for example in case of synthetic video).
12601 The pixel format name.
12604 The sample aspect ratio of the input frame, expressed in the form
12605 @var{num}/@var{den}.
12608 The size of the input frame. For the syntax of this option, check the
12609 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
12612 The type of interlaced mode ("P" for "progressive", "T" for top field first, "B"
12613 for bottom field first).
12616 This is 1 if the frame is a key frame, 0 otherwise.
12619 The picture type of the input frame ("I" for an I-frame, "P" for a
12620 P-frame, "B" for a B-frame, or "?" for an unknown type).
12621 Also refer to the documentation of the @code{AVPictureType} enum and of
12622 the @code{av_get_picture_type_char} function defined in
12623 @file{libavutil/avutil.h}.
12626 The Adler-32 checksum (printed in hexadecimal) of all the planes of the input frame.
12628 @item plane_checksum
12629 The Adler-32 checksum (printed in hexadecimal) of each plane of the input frame,
12630 expressed in the form "[@var{c0} @var{c1} @var{c2} @var{c3}]".
12633 @section showpalette
12635 Displays the 256 colors palette of each frame. This filter is only relevant for
12636 @var{pal8} pixel format frames.
12638 It accepts the following option:
12642 Set the size of the box used to represent one palette color entry. Default is
12643 @code{30} (for a @code{30x30} pixel box).
12646 @section shuffleframes
12648 Reorder and/or duplicate and/or drop video frames.
12650 It accepts the following parameters:
12654 Set the destination indexes of input frames.
12655 This is space or '|' separated list of indexes that maps input frames to output
12656 frames. Number of indexes also sets maximal value that each index may have.
12657 '-1' index have special meaning and that is to drop frame.
12660 The first frame has the index 0. The default is to keep the input unchanged.
12662 @subsection Examples
12666 Swap second and third frame of every three frames of the input:
12668 ffmpeg -i INPUT -vf "shuffleframes=0 2 1" OUTPUT
12672 Swap 10th and 1st frame of every ten frames of the input:
12674 ffmpeg -i INPUT -vf "shuffleframes=9 1 2 3 4 5 6 7 8 0" OUTPUT
12678 @section shuffleplanes
12680 Reorder and/or duplicate video planes.
12682 It accepts the following parameters:
12687 The index of the input plane to be used as the first output plane.
12690 The index of the input plane to be used as the second output plane.
12693 The index of the input plane to be used as the third output plane.
12696 The index of the input plane to be used as the fourth output plane.
12700 The first plane has the index 0. The default is to keep the input unchanged.
12702 @subsection Examples
12706 Swap the second and third planes of the input:
12708 ffmpeg -i INPUT -vf shuffleplanes=0:2:1:3 OUTPUT
12712 @anchor{signalstats}
12713 @section signalstats
12714 Evaluate various visual metrics that assist in determining issues associated
12715 with the digitization of analog video media.
12717 By default the filter will log these metadata values:
12721 Display the minimal Y value contained within the input frame. Expressed in
12725 Display the Y value at the 10% percentile within the input frame. Expressed in
12729 Display the average Y value within the input frame. Expressed in range of
12733 Display the Y value at the 90% percentile within the input frame. Expressed in
12737 Display the maximum Y value contained within the input frame. Expressed in
12741 Display the minimal U value contained within the input frame. Expressed in
12745 Display the U value at the 10% percentile within the input frame. Expressed in
12749 Display the average U value within the input frame. Expressed in range of
12753 Display the U value at the 90% percentile within the input frame. Expressed in
12757 Display the maximum U value contained within the input frame. Expressed in
12761 Display the minimal V value contained within the input frame. Expressed in
12765 Display the V value at the 10% percentile within the input frame. Expressed in
12769 Display the average V value within the input frame. Expressed in range of
12773 Display the V value at the 90% percentile within the input frame. Expressed in
12777 Display the maximum V value contained within the input frame. Expressed in
12781 Display the minimal saturation value contained within the input frame.
12782 Expressed in range of [0-~181.02].
12785 Display the saturation value at the 10% percentile within the input frame.
12786 Expressed in range of [0-~181.02].
12789 Display the average saturation value within the input frame. Expressed in range
12793 Display the saturation value at the 90% percentile within the input frame.
12794 Expressed in range of [0-~181.02].
12797 Display the maximum saturation value contained within the input frame.
12798 Expressed in range of [0-~181.02].
12801 Display the median value for hue within the input frame. Expressed in range of
12805 Display the average value for hue within the input frame. Expressed in range of
12809 Display the average of sample value difference between all values of the Y
12810 plane in the current frame and corresponding values of the previous input frame.
12811 Expressed in range of [0-255].
12814 Display the average of sample value difference between all values of the U
12815 plane in the current frame and corresponding values of the previous input frame.
12816 Expressed in range of [0-255].
12819 Display the average of sample value difference between all values of the V
12820 plane in the current frame and corresponding values of the previous input frame.
12821 Expressed in range of [0-255].
12824 Display bit depth of Y plane in current frame.
12825 Expressed in range of [0-16].
12828 Display bit depth of U plane in current frame.
12829 Expressed in range of [0-16].
12832 Display bit depth of V plane in current frame.
12833 Expressed in range of [0-16].
12836 The filter accepts the following options:
12842 @option{stat} specify an additional form of image analysis.
12843 @option{out} output video with the specified type of pixel highlighted.
12845 Both options accept the following values:
12849 Identify @var{temporal outliers} pixels. A @var{temporal outlier} is a pixel
12850 unlike the neighboring pixels of the same field. Examples of temporal outliers
12851 include the results of video dropouts, head clogs, or tape tracking issues.
12854 Identify @var{vertical line repetition}. Vertical line repetition includes
12855 similar rows of pixels within a frame. In born-digital video vertical line
12856 repetition is common, but this pattern is uncommon in video digitized from an
12857 analog source. When it occurs in video that results from the digitization of an
12858 analog source it can indicate concealment from a dropout compensator.
12861 Identify pixels that fall outside of legal broadcast range.
12865 Set the highlight color for the @option{out} option. The default color is
12869 @subsection Examples
12873 Output data of various video metrics:
12875 ffprobe -f lavfi movie=example.mov,signalstats="stat=tout+vrep+brng" -show_frames
12879 Output specific data about the minimum and maximum values of the Y plane per frame:
12881 ffprobe -f lavfi movie=example.mov,signalstats -show_entries frame_tags=lavfi.signalstats.YMAX,lavfi.signalstats.YMIN
12885 Playback video while highlighting pixels that are outside of broadcast range in red.
12887 ffplay example.mov -vf signalstats="out=brng:color=red"
12891 Playback video with signalstats metadata drawn over the frame.
12893 ffplay example.mov -vf signalstats=stat=brng+vrep+tout,drawtext=fontfile=FreeSerif.ttf:textfile=signalstat_drawtext.txt
12896 The contents of signalstat_drawtext.txt used in the command are:
12899 Y (%@{metadata:lavfi.signalstats.YMIN@}-%@{metadata:lavfi.signalstats.YMAX@})
12900 U (%@{metadata:lavfi.signalstats.UMIN@}-%@{metadata:lavfi.signalstats.UMAX@})
12901 V (%@{metadata:lavfi.signalstats.VMIN@}-%@{metadata:lavfi.signalstats.VMAX@})
12902 saturation maximum: %@{metadata:lavfi.signalstats.SATMAX@}
12910 Calculates the MPEG-7 Video Signature. The filter can handle more than one
12911 input. In this case the matching between the inputs can be calculated additionally.
12912 The filter always passes through the first input. The signature of each stream can
12913 be written into a file.
12915 It accepts the following options:
12919 Enable or disable the matching process.
12921 Available values are:
12925 Disable the calculation of a matching (default).
12927 Calculate the matching for the whole video and output whether the whole video
12928 matches or only parts.
12930 Calculate only until a matching is found or the video ends. Should be faster in
12935 Set the number of inputs. The option value must be a non negative integer.
12936 Default value is 1.
12939 Set the path to which the output is written. If there is more than one input,
12940 the path must be a prototype, i.e. must contain %d or %0nd (where n is a positive
12941 integer), that will be replaced with the input number. If no filename is
12942 specified, no output will be written. This is the default.
12945 Choose the output format.
12947 Available values are:
12951 Use the specified binary representation (default).
12953 Use the specified xml representation.
12957 Set threshold to detect one word as similar. The option value must be an integer
12958 greater than zero. The default value is 9000.
12961 Set threshold to detect all words as similar. The option value must be an integer
12962 greater than zero. The default value is 60000.
12965 Set threshold to detect frames as similar. The option value must be an integer
12966 greater than zero. The default value is 116.
12969 Set the minimum length of a sequence in frames to recognize it as matching
12970 sequence. The option value must be a non negative integer value.
12971 The default value is 0.
12974 Set the minimum relation, that matching frames to all frames must have.
12975 The option value must be a double value between 0 and 1. The default value is 0.5.
12978 @subsection Examples
12982 To calculate the signature of an input video and store it in signature.bin:
12984 ffmpeg -i input.mkv -vf signature=filename=signature.bin -map 0:v -f null -
12988 To detect whether two videos match and store the signatures in XML format in
12989 signature0.xml and signature1.xml:
12991 ffmpeg -i input1.mkv -i input2.mkv -filter_complex "[0:v][1:v] signature=nb_inputs=2:detectmode=full:format=xml:filename=signature%d.xml" -map :v -f null -
12999 Blur the input video without impacting the outlines.
13001 It accepts the following options:
13004 @item luma_radius, lr
13005 Set the luma radius. The option value must be a float number in
13006 the range [0.1,5.0] that specifies the variance of the gaussian filter
13007 used to blur the image (slower if larger). Default value is 1.0.
13009 @item luma_strength, ls
13010 Set the luma strength. The option value must be a float number
13011 in the range [-1.0,1.0] that configures the blurring. A value included
13012 in [0.0,1.0] will blur the image whereas a value included in
13013 [-1.0,0.0] will sharpen the image. Default value is 1.0.
13015 @item luma_threshold, lt
13016 Set the luma threshold used as a coefficient to determine
13017 whether a pixel should be blurred or not. The option value must be an
13018 integer in the range [-30,30]. A value of 0 will filter all the image,
13019 a value included in [0,30] will filter flat areas and a value included
13020 in [-30,0] will filter edges. Default value is 0.
13022 @item chroma_radius, cr
13023 Set the chroma radius. The option value must be a float number in
13024 the range [0.1,5.0] that specifies the variance of the gaussian filter
13025 used to blur the image (slower if larger). Default value is @option{luma_radius}.
13027 @item chroma_strength, cs
13028 Set the chroma strength. The option value must be a float number
13029 in the range [-1.0,1.0] that configures the blurring. A value included
13030 in [0.0,1.0] will blur the image whereas a value included in
13031 [-1.0,0.0] will sharpen the image. Default value is @option{luma_strength}.
13033 @item chroma_threshold, ct
13034 Set the chroma threshold used as a coefficient to determine
13035 whether a pixel should be blurred or not. The option value must be an
13036 integer in the range [-30,30]. A value of 0 will filter all the image,
13037 a value included in [0,30] will filter flat areas and a value included
13038 in [-30,0] will filter edges. Default value is @option{luma_threshold}.
13041 If a chroma option is not explicitly set, the corresponding luma value
13046 Obtain the SSIM (Structural SImilarity Metric) between two input videos.
13048 This filter takes in input two input videos, the first input is
13049 considered the "main" source and is passed unchanged to the
13050 output. The second input is used as a "reference" video for computing
13053 Both video inputs must have the same resolution and pixel format for
13054 this filter to work correctly. Also it assumes that both inputs
13055 have the same number of frames, which are compared one by one.
13057 The filter stores the calculated SSIM of each frame.
13059 The description of the accepted parameters follows.
13062 @item stats_file, f
13063 If specified the filter will use the named file to save the SSIM of
13064 each individual frame. When filename equals "-" the data is sent to
13068 The file printed if @var{stats_file} is selected, contains a sequence of
13069 key/value pairs of the form @var{key}:@var{value} for each compared
13072 A description of each shown parameter follows:
13076 sequential number of the input frame, starting from 1
13078 @item Y, U, V, R, G, B
13079 SSIM of the compared frames for the component specified by the suffix.
13082 SSIM of the compared frames for the whole frame.
13085 Same as above but in dB representation.
13090 movie=ref_movie.mpg, setpts=PTS-STARTPTS [main];
13091 [main][ref] ssim="stats_file=stats.log" [out]
13094 On this example the input file being processed is compared with the
13095 reference file @file{ref_movie.mpg}. The SSIM of each individual frame
13096 is stored in @file{stats.log}.
13098 Another example with both psnr and ssim at same time:
13100 ffmpeg -i main.mpg -i ref.mpg -lavfi "ssim;[0:v][1:v]psnr" -f null -
13105 Convert between different stereoscopic image formats.
13107 The filters accept the following options:
13111 Set stereoscopic image format of input.
13113 Available values for input image formats are:
13116 side by side parallel (left eye left, right eye right)
13119 side by side crosseye (right eye left, left eye right)
13122 side by side parallel with half width resolution
13123 (left eye left, right eye right)
13126 side by side crosseye with half width resolution
13127 (right eye left, left eye right)
13130 above-below (left eye above, right eye below)
13133 above-below (right eye above, left eye below)
13136 above-below with half height resolution
13137 (left eye above, right eye below)
13140 above-below with half height resolution
13141 (right eye above, left eye below)
13144 alternating frames (left eye first, right eye second)
13147 alternating frames (right eye first, left eye second)
13150 interleaved rows (left eye has top row, right eye starts on next row)
13153 interleaved rows (right eye has top row, left eye starts on next row)
13156 interleaved columns, left eye first
13159 interleaved columns, right eye first
13161 Default value is @samp{sbsl}.
13165 Set stereoscopic image format of output.
13169 side by side parallel (left eye left, right eye right)
13172 side by side crosseye (right eye left, left eye right)
13175 side by side parallel with half width resolution
13176 (left eye left, right eye right)
13179 side by side crosseye with half width resolution
13180 (right eye left, left eye right)
13183 above-below (left eye above, right eye below)
13186 above-below (right eye above, left eye below)
13189 above-below with half height resolution
13190 (left eye above, right eye below)
13193 above-below with half height resolution
13194 (right eye above, left eye below)
13197 alternating frames (left eye first, right eye second)
13200 alternating frames (right eye first, left eye second)
13203 interleaved rows (left eye has top row, right eye starts on next row)
13206 interleaved rows (right eye has top row, left eye starts on next row)
13209 anaglyph red/blue gray
13210 (red filter on left eye, blue filter on right eye)
13213 anaglyph red/green gray
13214 (red filter on left eye, green filter on right eye)
13217 anaglyph red/cyan gray
13218 (red filter on left eye, cyan filter on right eye)
13221 anaglyph red/cyan half colored
13222 (red filter on left eye, cyan filter on right eye)
13225 anaglyph red/cyan color
13226 (red filter on left eye, cyan filter on right eye)
13229 anaglyph red/cyan color optimized with the least squares projection of dubois
13230 (red filter on left eye, cyan filter on right eye)
13233 anaglyph green/magenta gray
13234 (green filter on left eye, magenta filter on right eye)
13237 anaglyph green/magenta half colored
13238 (green filter on left eye, magenta filter on right eye)
13241 anaglyph green/magenta colored
13242 (green filter on left eye, magenta filter on right eye)
13245 anaglyph green/magenta color optimized with the least squares projection of dubois
13246 (green filter on left eye, magenta filter on right eye)
13249 anaglyph yellow/blue gray
13250 (yellow filter on left eye, blue filter on right eye)
13253 anaglyph yellow/blue half colored
13254 (yellow filter on left eye, blue filter on right eye)
13257 anaglyph yellow/blue colored
13258 (yellow filter on left eye, blue filter on right eye)
13261 anaglyph yellow/blue color optimized with the least squares projection of dubois
13262 (yellow filter on left eye, blue filter on right eye)
13265 mono output (left eye only)
13268 mono output (right eye only)
13271 checkerboard, left eye first
13274 checkerboard, right eye first
13277 interleaved columns, left eye first
13280 interleaved columns, right eye first
13286 Default value is @samp{arcd}.
13289 @subsection Examples
13293 Convert input video from side by side parallel to anaglyph yellow/blue dubois:
13299 Convert input video from above below (left eye above, right eye below) to side by side crosseye.
13305 @section streamselect, astreamselect
13306 Select video or audio streams.
13308 The filter accepts the following options:
13312 Set number of inputs. Default is 2.
13315 Set input indexes to remap to outputs.
13318 @subsection Commands
13320 The @code{streamselect} and @code{astreamselect} filter supports the following
13325 Set input indexes to remap to outputs.
13328 @subsection Examples
13332 Select first 5 seconds 1st stream and rest of time 2nd stream:
13334 sendcmd='5.0 streamselect map 1',streamselect=inputs=2:map=0
13338 Same as above, but for audio:
13340 asendcmd='5.0 astreamselect map 1',astreamselect=inputs=2:map=0
13345 Apply sobel operator to input video stream.
13347 The filter accepts the following option:
13351 Set which planes will be processed, unprocessed planes will be copied.
13352 By default value 0xf, all planes will be processed.
13355 Set value which will be multiplied with filtered result.
13358 Set value which will be added to filtered result.
13364 Apply a simple postprocessing filter that compresses and decompresses the image
13365 at several (or - in the case of @option{quality} level @code{6} - all) shifts
13366 and average the results.
13368 The filter accepts the following options:
13372 Set quality. This option defines the number of levels for averaging. It accepts
13373 an integer in the range 0-6. If set to @code{0}, the filter will have no
13374 effect. A value of @code{6} means the higher quality. For each increment of
13375 that value the speed drops by a factor of approximately 2. Default value is
13379 Force a constant quantization parameter. If not set, the filter will use the QP
13380 from the video stream (if available).
13383 Set thresholding mode. Available modes are:
13387 Set hard thresholding (default).
13389 Set soft thresholding (better de-ringing effect, but likely blurrier).
13392 @item use_bframe_qp
13393 Enable the use of the QP from the B-Frames if set to @code{1}. Using this
13394 option may cause flicker since the B-Frames have often larger QP. Default is
13395 @code{0} (not enabled).
13401 Draw subtitles on top of input video using the libass library.
13403 To enable compilation of this filter you need to configure FFmpeg with
13404 @code{--enable-libass}. This filter also requires a build with libavcodec and
13405 libavformat to convert the passed subtitles file to ASS (Advanced Substation
13406 Alpha) subtitles format.
13408 The filter accepts the following options:
13412 Set the filename of the subtitle file to read. It must be specified.
13414 @item original_size
13415 Specify the size of the original video, the video for which the ASS file
13416 was composed. For the syntax of this option, check the
13417 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13418 Due to a misdesign in ASS aspect ratio arithmetic, this is necessary to
13419 correctly scale the fonts if the aspect ratio has been changed.
13422 Set a directory path containing fonts that can be used by the filter.
13423 These fonts will be used in addition to whatever the font provider uses.
13426 Set subtitles input character encoding. @code{subtitles} filter only. Only
13427 useful if not UTF-8.
13429 @item stream_index, si
13430 Set subtitles stream index. @code{subtitles} filter only.
13433 Override default style or script info parameters of the subtitles. It accepts a
13434 string containing ASS style format @code{KEY=VALUE} couples separated by ",".
13437 If the first key is not specified, it is assumed that the first value
13438 specifies the @option{filename}.
13440 For example, to render the file @file{sub.srt} on top of the input
13441 video, use the command:
13446 which is equivalent to:
13448 subtitles=filename=sub.srt
13451 To render the default subtitles stream from file @file{video.mkv}, use:
13453 subtitles=video.mkv
13456 To render the second subtitles stream from that file, use:
13458 subtitles=video.mkv:si=1
13461 To make the subtitles stream from @file{sub.srt} appear in transparent green
13462 @code{DejaVu Serif}, use:
13464 subtitles=sub.srt:force_style='FontName=DejaVu Serif,PrimaryColour=&HAA00FF00'
13467 @section super2xsai
13469 Scale the input by 2x and smooth using the Super2xSaI (Scale and
13470 Interpolate) pixel art scaling algorithm.
13472 Useful for enlarging pixel art images without reducing sharpness.
13476 Swap two rectangular objects in video.
13478 This filter accepts the following options:
13488 Set 1st rect x coordinate.
13491 Set 1st rect y coordinate.
13494 Set 2nd rect x coordinate.
13497 Set 2nd rect y coordinate.
13499 All expressions are evaluated once for each frame.
13502 The all options are expressions containing the following constants:
13507 The input width and height.
13510 same as @var{w} / @var{h}
13513 input sample aspect ratio
13516 input display aspect ratio, it is the same as (@var{w} / @var{h}) * @var{sar}
13519 The number of the input frame, starting from 0.
13522 The timestamp expressed in seconds. It's NAN if the input timestamp is unknown.
13525 the position in the file of the input frame, NAN if unknown
13533 Apply telecine process to the video.
13535 This filter accepts the following options:
13544 The default value is @code{top}.
13548 A string of numbers representing the pulldown pattern you wish to apply.
13549 The default value is @code{23}.
13553 Some typical patterns:
13558 24p: 2332 (preferred)
13565 24p: 222222222223 ("Euro pulldown")
13572 Apply threshold effect to video stream.
13574 This filter needs four video streams to perform thresholding.
13575 First stream is stream we are filtering.
13576 Second stream is holding threshold values, third stream is holding min values,
13577 and last, fourth stream is holding max values.
13579 The filter accepts the following option:
13583 Set which planes will be processed, unprocessed planes will be copied.
13584 By default value 0xf, all planes will be processed.
13587 For example if first stream pixel's component value is less then threshold value
13588 of pixel component from 2nd threshold stream, third stream value will picked,
13589 otherwise fourth stream pixel component value will be picked.
13591 Using color source filter one can perform various types of thresholding:
13593 @subsection Examples
13597 Binary threshold, using gray color as threshold:
13599 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=black -f lavfi -i color=white -lavfi threshold output.avi
13603 Inverted binary threshold, using gray color as threshold:
13605 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -f lavfi -i color=black -lavfi threshold output.avi
13609 Truncate binary threshold, using gray color as threshold:
13611 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=gray -lavfi threshold output.avi
13615 Threshold to zero, using gray color as threshold:
13617 ffmpeg -i 320x240.avi -f lavfi -i color=gray -f lavfi -i color=white -i 320x240.avi -lavfi threshold output.avi
13621 Inverted threshold to zero, using gray color as threshold:
13623 ffmpeg -i 320x240.avi -f lavfi -i color=gray -i 320x240.avi -f lavfi -i color=white -lavfi threshold output.avi
13628 Select the most representative frame in a given sequence of consecutive frames.
13630 The filter accepts the following options:
13634 Set the frames batch size to analyze; in a set of @var{n} frames, the filter
13635 will pick one of them, and then handle the next batch of @var{n} frames until
13636 the end. Default is @code{100}.
13639 Since the filter keeps track of the whole frames sequence, a bigger @var{n}
13640 value will result in a higher memory usage, so a high value is not recommended.
13642 @subsection Examples
13646 Extract one picture each 50 frames:
13652 Complete example of a thumbnail creation with @command{ffmpeg}:
13654 ffmpeg -i in.avi -vf thumbnail,scale=300:200 -frames:v 1 out.png
13660 Tile several successive frames together.
13662 The filter accepts the following options:
13667 Set the grid size (i.e. the number of lines and columns). For the syntax of
13668 this option, check the
13669 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
13672 Set the maximum number of frames to render in the given area. It must be less
13673 than or equal to @var{w}x@var{h}. The default value is @code{0}, meaning all
13674 the area will be used.
13677 Set the outer border margin in pixels.
13680 Set the inner border thickness (i.e. the number of pixels between frames). For
13681 more advanced padding options (such as having different values for the edges),
13682 refer to the pad video filter.
13685 Specify the color of the unused area. For the syntax of this option, check the
13686 "Color" section in the ffmpeg-utils manual. The default value of @var{color}
13690 @subsection Examples
13694 Produce 8x8 PNG tiles of all keyframes (@option{-skip_frame nokey}) in a movie:
13696 ffmpeg -skip_frame nokey -i file.avi -vf 'scale=128:72,tile=8x8' -an -vsync 0 keyframes%03d.png
13698 The @option{-vsync 0} is necessary to prevent @command{ffmpeg} from
13699 duplicating each output frame to accommodate the originally detected frame
13703 Display @code{5} pictures in an area of @code{3x2} frames,
13704 with @code{7} pixels between them, and @code{2} pixels of initial margin, using
13705 mixed flat and named options:
13707 tile=3x2:nb_frames=5:padding=7:margin=2
13711 @section tinterlace
13713 Perform various types of temporal field interlacing.
13715 Frames are counted starting from 1, so the first input frame is
13718 The filter accepts the following options:
13723 Specify the mode of the interlacing. This option can also be specified
13724 as a value alone. See below for a list of values for this option.
13726 Available values are:
13730 Move odd frames into the upper field, even into the lower field,
13731 generating a double height frame at half frame rate.
13735 Frame 1 Frame 2 Frame 3 Frame 4
13737 11111 22222 33333 44444
13738 11111 22222 33333 44444
13739 11111 22222 33333 44444
13740 11111 22222 33333 44444
13754 Only output odd frames, even frames are dropped, generating a frame with
13755 unchanged height at half frame rate.
13760 Frame 1 Frame 2 Frame 3 Frame 4
13762 11111 22222 33333 44444
13763 11111 22222 33333 44444
13764 11111 22222 33333 44444
13765 11111 22222 33333 44444
13775 Only output even frames, odd frames are dropped, generating a frame with
13776 unchanged height at half frame rate.
13781 Frame 1 Frame 2 Frame 3 Frame 4
13783 11111 22222 33333 44444
13784 11111 22222 33333 44444
13785 11111 22222 33333 44444
13786 11111 22222 33333 44444
13796 Expand each frame to full height, but pad alternate lines with black,
13797 generating a frame with double height at the same input frame rate.
13802 Frame 1 Frame 2 Frame 3 Frame 4
13804 11111 22222 33333 44444
13805 11111 22222 33333 44444
13806 11111 22222 33333 44444
13807 11111 22222 33333 44444
13810 11111 ..... 33333 .....
13811 ..... 22222 ..... 44444
13812 11111 ..... 33333 .....
13813 ..... 22222 ..... 44444
13814 11111 ..... 33333 .....
13815 ..... 22222 ..... 44444
13816 11111 ..... 33333 .....
13817 ..... 22222 ..... 44444
13821 @item interleave_top, 4
13822 Interleave the upper field from odd frames with the lower field from
13823 even frames, generating a frame with unchanged height at half frame rate.
13828 Frame 1 Frame 2 Frame 3 Frame 4
13830 11111<- 22222 33333<- 44444
13831 11111 22222<- 33333 44444<-
13832 11111<- 22222 33333<- 44444
13833 11111 22222<- 33333 44444<-
13843 @item interleave_bottom, 5
13844 Interleave the lower field from odd frames with the upper field from
13845 even frames, generating a frame with unchanged height at half frame rate.
13850 Frame 1 Frame 2 Frame 3 Frame 4
13852 11111 22222<- 33333 44444<-
13853 11111<- 22222 33333<- 44444
13854 11111 22222<- 33333 44444<-
13855 11111<- 22222 33333<- 44444
13865 @item interlacex2, 6
13866 Double frame rate with unchanged height. Frames are inserted each
13867 containing the second temporal field from the previous input frame and
13868 the first temporal field from the next input frame. This mode relies on
13869 the top_field_first flag. Useful for interlaced video displays with no
13870 field synchronisation.
13875 Frame 1 Frame 2 Frame 3 Frame 4
13877 11111 22222 33333 44444
13878 11111 22222 33333 44444
13879 11111 22222 33333 44444
13880 11111 22222 33333 44444
13883 11111 22222 22222 33333 33333 44444 44444
13884 11111 11111 22222 22222 33333 33333 44444
13885 11111 22222 22222 33333 33333 44444 44444
13886 11111 11111 22222 22222 33333 33333 44444
13891 Move odd frames into the upper field, even into the lower field,
13892 generating a double height frame at same frame rate.
13897 Frame 1 Frame 2 Frame 3 Frame 4
13899 11111 22222 33333 44444
13900 11111 22222 33333 44444
13901 11111 22222 33333 44444
13902 11111 22222 33333 44444
13905 11111 33333 33333 55555
13906 22222 22222 44444 44444
13907 11111 33333 33333 55555
13908 22222 22222 44444 44444
13909 11111 33333 33333 55555
13910 22222 22222 44444 44444
13911 11111 33333 33333 55555
13912 22222 22222 44444 44444
13917 Numeric values are deprecated but are accepted for backward
13918 compatibility reasons.
13920 Default mode is @code{merge}.
13923 Specify flags influencing the filter process.
13925 Available value for @var{flags} is:
13928 @item low_pass_filter, vlfp
13929 Enable linear vertical low-pass filtering in the filter.
13930 Vertical low-pass filtering is required when creating an interlaced
13931 destination from a progressive source which contains high-frequency
13932 vertical detail. Filtering will reduce interlace 'twitter' and Moire
13935 @item complex_filter, cvlfp
13936 Enable complex vertical low-pass filtering.
13937 This will slightly less reduce interlace 'twitter' and Moire
13938 patterning but better retain detail and subjective sharpness impression.
13942 Vertical low-pass filtering can only be enabled for @option{mode}
13943 @var{interleave_top} and @var{interleave_bottom}.
13949 Transpose rows with columns in the input video and optionally flip it.
13951 It accepts the following parameters:
13956 Specify the transposition direction.
13958 Can assume the following values:
13960 @item 0, 4, cclock_flip
13961 Rotate by 90 degrees counterclockwise and vertically flip (default), that is:
13969 Rotate by 90 degrees clockwise, that is:
13977 Rotate by 90 degrees counterclockwise, that is:
13984 @item 3, 7, clock_flip
13985 Rotate by 90 degrees clockwise and vertically flip, that is:
13993 For values between 4-7, the transposition is only done if the input
13994 video geometry is portrait and not landscape. These values are
13995 deprecated, the @code{passthrough} option should be used instead.
13997 Numerical values are deprecated, and should be dropped in favor of
13998 symbolic constants.
14001 Do not apply the transposition if the input geometry matches the one
14002 specified by the specified value. It accepts the following values:
14005 Always apply transposition.
14007 Preserve portrait geometry (when @var{height} >= @var{width}).
14009 Preserve landscape geometry (when @var{width} >= @var{height}).
14012 Default value is @code{none}.
14015 For example to rotate by 90 degrees clockwise and preserve portrait
14018 transpose=dir=1:passthrough=portrait
14021 The command above can also be specified as:
14023 transpose=1:portrait
14027 Trim the input so that the output contains one continuous subpart of the input.
14029 It accepts the following parameters:
14032 Specify the time of the start of the kept section, i.e. the frame with the
14033 timestamp @var{start} will be the first frame in the output.
14036 Specify the time of the first frame that will be dropped, i.e. the frame
14037 immediately preceding the one with the timestamp @var{end} will be the last
14038 frame in the output.
14041 This is the same as @var{start}, except this option sets the start timestamp
14042 in timebase units instead of seconds.
14045 This is the same as @var{end}, except this option sets the end timestamp
14046 in timebase units instead of seconds.
14049 The maximum duration of the output in seconds.
14052 The number of the first frame that should be passed to the output.
14055 The number of the first frame that should be dropped.
14058 @option{start}, @option{end}, and @option{duration} are expressed as time
14059 duration specifications; see
14060 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
14061 for the accepted syntax.
14063 Note that the first two sets of the start/end options and the @option{duration}
14064 option look at the frame timestamp, while the _frame variants simply count the
14065 frames that pass through the filter. Also note that this filter does not modify
14066 the timestamps. If you wish for the output timestamps to start at zero, insert a
14067 setpts filter after the trim filter.
14069 If multiple start or end options are set, this filter tries to be greedy and
14070 keep all the frames that match at least one of the specified constraints. To keep
14071 only the part that matches all the constraints at once, chain multiple trim
14074 The defaults are such that all the input is kept. So it is possible to set e.g.
14075 just the end values to keep everything before the specified time.
14080 Drop everything except the second minute of input:
14082 ffmpeg -i INPUT -vf trim=60:120
14086 Keep only the first second:
14088 ffmpeg -i INPUT -vf trim=duration=1
14097 Sharpen or blur the input video.
14099 It accepts the following parameters:
14102 @item luma_msize_x, lx
14103 Set the luma matrix horizontal size. It must be an odd integer between
14104 3 and 23. The default value is 5.
14106 @item luma_msize_y, ly
14107 Set the luma matrix vertical size. It must be an odd integer between 3
14108 and 23. The default value is 5.
14110 @item luma_amount, la
14111 Set the luma effect strength. It must be a floating point number, reasonable
14112 values lay between -1.5 and 1.5.
14114 Negative values will blur the input video, while positive values will
14115 sharpen it, a value of zero will disable the effect.
14117 Default value is 1.0.
14119 @item chroma_msize_x, cx
14120 Set the chroma matrix horizontal size. It must be an odd integer
14121 between 3 and 23. The default value is 5.
14123 @item chroma_msize_y, cy
14124 Set the chroma matrix vertical size. It must be an odd integer
14125 between 3 and 23. The default value is 5.
14127 @item chroma_amount, ca
14128 Set the chroma effect strength. It must be a floating point number, reasonable
14129 values lay between -1.5 and 1.5.
14131 Negative values will blur the input video, while positive values will
14132 sharpen it, a value of zero will disable the effect.
14134 Default value is 0.0.
14137 If set to 1, specify using OpenCL capabilities, only available if
14138 FFmpeg was configured with @code{--enable-opencl}. Default value is 0.
14142 All parameters are optional and default to the equivalent of the
14143 string '5:5:1.0:5:5:0.0'.
14145 @subsection Examples
14149 Apply strong luma sharpen effect:
14151 unsharp=luma_msize_x=7:luma_msize_y=7:luma_amount=2.5
14155 Apply a strong blur of both luma and chroma parameters:
14157 unsharp=7:7:-2:7:7:-2
14163 Apply ultra slow/simple postprocessing filter that compresses and decompresses
14164 the image at several (or - in the case of @option{quality} level @code{8} - all)
14165 shifts and average the results.
14167 The way this differs from the behavior of spp is that uspp actually encodes &
14168 decodes each case with libavcodec Snow, whereas spp uses a simplified intra only 8x8
14169 DCT similar to MJPEG.
14171 The filter accepts the following options:
14175 Set quality. This option defines the number of levels for averaging. It accepts
14176 an integer in the range 0-8. If set to @code{0}, the filter will have no
14177 effect. A value of @code{8} means the higher quality. For each increment of
14178 that value the speed drops by a factor of approximately 2. Default value is
14182 Force a constant quantization parameter. If not set, the filter will use the QP
14183 from the video stream (if available).
14186 @section vaguedenoiser
14188 Apply a wavelet based denoiser.
14190 It transforms each frame from the video input into the wavelet domain,
14191 using Cohen-Daubechies-Feauveau 9/7. Then it applies some filtering to
14192 the obtained coefficients. It does an inverse wavelet transform after.
14193 Due to wavelet properties, it should give a nice smoothed result, and
14194 reduced noise, without blurring picture features.
14196 This filter accepts the following options:
14200 The filtering strength. The higher, the more filtered the video will be.
14201 Hard thresholding can use a higher threshold than soft thresholding
14202 before the video looks overfiltered.
14205 The filtering method the filter will use.
14207 It accepts the following values:
14210 All values under the threshold will be zeroed.
14213 All values under the threshold will be zeroed. All values above will be
14214 reduced by the threshold.
14217 Scales or nullifies coefficients - intermediary between (more) soft and
14218 (less) hard thresholding.
14222 Number of times, the wavelet will decompose the picture. Picture can't
14223 be decomposed beyond a particular point (typically, 8 for a 640x480
14224 frame - as 2^9 = 512 > 480)
14227 Partial of full denoising (limited coefficients shrinking), from 0 to 100.
14230 A list of the planes to process. By default all planes are processed.
14233 @section vectorscope
14235 Display 2 color component values in the two dimensional graph (which is called
14238 This filter accepts the following options:
14242 Set vectorscope mode.
14244 It accepts the following values:
14247 Gray values are displayed on graph, higher brightness means more pixels have
14248 same component color value on location in graph. This is the default mode.
14251 Gray values are displayed on graph. Surrounding pixels values which are not
14252 present in video frame are drawn in gradient of 2 color components which are
14253 set by option @code{x} and @code{y}. The 3rd color component is static.
14256 Actual color components values present in video frame are displayed on graph.
14259 Similar as color2 but higher frequency of same values @code{x} and @code{y}
14260 on graph increases value of another color component, which is luminance by
14261 default values of @code{x} and @code{y}.
14264 Actual colors present in video frame are displayed on graph. If two different
14265 colors map to same position on graph then color with higher value of component
14266 not present in graph is picked.
14269 Gray values are displayed on graph. Similar to @code{color} but with 3rd color
14270 component picked from radial gradient.
14274 Set which color component will be represented on X-axis. Default is @code{1}.
14277 Set which color component will be represented on Y-axis. Default is @code{2}.
14280 Set intensity, used by modes: gray, color, color3 and color5 for increasing brightness
14281 of color component which represents frequency of (X, Y) location in graph.
14286 No envelope, this is default.
14289 Instant envelope, even darkest single pixel will be clearly highlighted.
14292 Hold maximum and minimum values presented in graph over time. This way you
14293 can still spot out of range values without constantly looking at vectorscope.
14296 Peak and instant envelope combined together.
14300 Set what kind of graticule to draw.
14308 Set graticule opacity.
14311 Set graticule flags.
14315 Draw graticule for white point.
14318 Draw graticule for black point.
14321 Draw color points short names.
14325 Set background opacity.
14327 @item lthreshold, l
14328 Set low threshold for color component not represented on X or Y axis.
14329 Values lower than this value will be ignored. Default is 0.
14330 Note this value is multiplied with actual max possible value one pixel component
14331 can have. So for 8-bit input and low threshold value of 0.1 actual threshold
14334 @item hthreshold, h
14335 Set high threshold for color component not represented on X or Y axis.
14336 Values higher than this value will be ignored. Default is 1.
14337 Note this value is multiplied with actual max possible value one pixel component
14338 can have. So for 8-bit input and high threshold value of 0.9 actual threshold
14339 is 0.9 * 255 = 230.
14341 @item colorspace, c
14342 Set what kind of colorspace to use when drawing graticule.
14351 @anchor{vidstabdetect}
14352 @section vidstabdetect
14354 Analyze video stabilization/deshaking. Perform pass 1 of 2, see
14355 @ref{vidstabtransform} for pass 2.
14357 This filter generates a file with relative translation and rotation
14358 transform information about subsequent frames, which is then used by
14359 the @ref{vidstabtransform} filter.
14361 To enable compilation of this filter you need to configure FFmpeg with
14362 @code{--enable-libvidstab}.
14364 This filter accepts the following options:
14368 Set the path to the file used to write the transforms information.
14369 Default value is @file{transforms.trf}.
14372 Set how shaky the video is and how quick the camera is. It accepts an
14373 integer in the range 1-10, a value of 1 means little shakiness, a
14374 value of 10 means strong shakiness. Default value is 5.
14377 Set the accuracy of the detection process. It must be a value in the
14378 range 1-15. A value of 1 means low accuracy, a value of 15 means high
14379 accuracy. Default value is 15.
14382 Set stepsize of the search process. The region around minimum is
14383 scanned with 1 pixel resolution. Default value is 6.
14386 Set minimum contrast. Below this value a local measurement field is
14387 discarded. Must be a floating point value in the range 0-1. Default
14391 Set reference frame number for tripod mode.
14393 If enabled, the motion of the frames is compared to a reference frame
14394 in the filtered stream, identified by the specified number. The idea
14395 is to compensate all movements in a more-or-less static scene and keep
14396 the camera view absolutely still.
14398 If set to 0, it is disabled. The frames are counted starting from 1.
14401 Show fields and transforms in the resulting frames. It accepts an
14402 integer in the range 0-2. Default value is 0, which disables any
14406 @subsection Examples
14410 Use default values:
14416 Analyze strongly shaky movie and put the results in file
14417 @file{mytransforms.trf}:
14419 vidstabdetect=shakiness=10:accuracy=15:result="mytransforms.trf"
14423 Visualize the result of internal transformations in the resulting
14426 vidstabdetect=show=1
14430 Analyze a video with medium shakiness using @command{ffmpeg}:
14432 ffmpeg -i input -vf vidstabdetect=shakiness=5:show=1 dummy.avi
14436 @anchor{vidstabtransform}
14437 @section vidstabtransform
14439 Video stabilization/deshaking: pass 2 of 2,
14440 see @ref{vidstabdetect} for pass 1.
14442 Read a file with transform information for each frame and
14443 apply/compensate them. Together with the @ref{vidstabdetect}
14444 filter this can be used to deshake videos. See also
14445 @url{http://public.hronopik.de/vid.stab}. It is important to also use
14446 the @ref{unsharp} filter, see below.
14448 To enable compilation of this filter you need to configure FFmpeg with
14449 @code{--enable-libvidstab}.
14451 @subsection Options
14455 Set path to the file used to read the transforms. Default value is
14456 @file{transforms.trf}.
14459 Set the number of frames (value*2 + 1) used for lowpass filtering the
14460 camera movements. Default value is 10.
14462 For example a number of 10 means that 21 frames are used (10 in the
14463 past and 10 in the future) to smoothen the motion in the video. A
14464 larger value leads to a smoother video, but limits the acceleration of
14465 the camera (pan/tilt movements). 0 is a special case where a static
14466 camera is simulated.
14469 Set the camera path optimization algorithm.
14471 Accepted values are:
14474 gaussian kernel low-pass filter on camera motion (default)
14476 averaging on transformations
14480 Set maximal number of pixels to translate frames. Default value is -1,
14484 Set maximal angle in radians (degree*PI/180) to rotate frames. Default
14485 value is -1, meaning no limit.
14488 Specify how to deal with borders that may be visible due to movement
14491 Available values are:
14494 keep image information from previous frame (default)
14496 fill the border black
14500 Invert transforms if set to 1. Default value is 0.
14503 Consider transforms as relative to previous frame if set to 1,
14504 absolute if set to 0. Default value is 0.
14507 Set percentage to zoom. A positive value will result in a zoom-in
14508 effect, a negative value in a zoom-out effect. Default value is 0 (no
14512 Set optimal zooming to avoid borders.
14514 Accepted values are:
14519 optimal static zoom value is determined (only very strong movements
14520 will lead to visible borders) (default)
14522 optimal adaptive zoom value is determined (no borders will be
14523 visible), see @option{zoomspeed}
14526 Note that the value given at zoom is added to the one calculated here.
14529 Set percent to zoom maximally each frame (enabled when
14530 @option{optzoom} is set to 2). Range is from 0 to 5, default value is
14534 Specify type of interpolation.
14536 Available values are:
14541 linear only horizontal
14543 linear in both directions (default)
14545 cubic in both directions (slow)
14549 Enable virtual tripod mode if set to 1, which is equivalent to
14550 @code{relative=0:smoothing=0}. Default value is 0.
14552 Use also @code{tripod} option of @ref{vidstabdetect}.
14555 Increase log verbosity if set to 1. Also the detected global motions
14556 are written to the temporary file @file{global_motions.trf}. Default
14560 @subsection Examples
14564 Use @command{ffmpeg} for a typical stabilization with default values:
14566 ffmpeg -i inp.mpeg -vf vidstabtransform,unsharp=5:5:0.8:3:3:0.4 inp_stabilized.mpeg
14569 Note the use of the @ref{unsharp} filter which is always recommended.
14572 Zoom in a bit more and load transform data from a given file:
14574 vidstabtransform=zoom=5:input="mytransforms.trf"
14578 Smoothen the video even more:
14580 vidstabtransform=smoothing=30
14586 Flip the input video vertically.
14588 For example, to vertically flip a video with @command{ffmpeg}:
14590 ffmpeg -i in.avi -vf "vflip" out.avi
14596 Make or reverse a natural vignetting effect.
14598 The filter accepts the following options:
14602 Set lens angle expression as a number of radians.
14604 The value is clipped in the @code{[0,PI/2]} range.
14606 Default value: @code{"PI/5"}
14610 Set center coordinates expressions. Respectively @code{"w/2"} and @code{"h/2"}
14614 Set forward/backward mode.
14616 Available modes are:
14619 The larger the distance from the central point, the darker the image becomes.
14622 The larger the distance from the central point, the brighter the image becomes.
14623 This can be used to reverse a vignette effect, though there is no automatic
14624 detection to extract the lens @option{angle} and other settings (yet). It can
14625 also be used to create a burning effect.
14628 Default value is @samp{forward}.
14631 Set evaluation mode for the expressions (@option{angle}, @option{x0}, @option{y0}).
14633 It accepts the following values:
14636 Evaluate expressions only once during the filter initialization.
14639 Evaluate expressions for each incoming frame. This is way slower than the
14640 @samp{init} mode since it requires all the scalers to be re-computed, but it
14641 allows advanced dynamic expressions.
14644 Default value is @samp{init}.
14647 Set dithering to reduce the circular banding effects. Default is @code{1}
14651 Set vignette aspect. This setting allows one to adjust the shape of the vignette.
14652 Setting this value to the SAR of the input will make a rectangular vignetting
14653 following the dimensions of the video.
14655 Default is @code{1/1}.
14658 @subsection Expressions
14660 The @option{alpha}, @option{x0} and @option{y0} expressions can contain the
14661 following parameters.
14666 input width and height
14669 the number of input frame, starting from 0
14672 the PTS (Presentation TimeStamp) time of the filtered video frame, expressed in
14673 @var{TB} units, NAN if undefined
14676 frame rate of the input video, NAN if the input frame rate is unknown
14679 the PTS (Presentation TimeStamp) of the filtered video frame,
14680 expressed in seconds, NAN if undefined
14683 time base of the input video
14687 @subsection Examples
14691 Apply simple strong vignetting effect:
14697 Make a flickering vignetting:
14699 vignette='PI/4+random(1)*PI/50':eval=frame
14705 Stack input videos vertically.
14707 All streams must be of same pixel format and of same width.
14709 Note that this filter is faster than using @ref{overlay} and @ref{pad} filter
14710 to create same output.
14712 The filter accept the following option:
14716 Set number of input streams. Default is 2.
14719 If set to 1, force the output to terminate when the shortest input
14720 terminates. Default value is 0.
14725 Deinterlace the input video ("w3fdif" stands for "Weston 3 Field
14726 Deinterlacing Filter").
14728 Based on the process described by Martin Weston for BBC R&D, and
14729 implemented based on the de-interlace algorithm written by Jim
14730 Easterbrook for BBC R&D, the Weston 3 field deinterlacing filter
14731 uses filter coefficients calculated by BBC R&D.
14733 There are two sets of filter coefficients, so called "simple":
14734 and "complex". Which set of filter coefficients is used can
14735 be set by passing an optional parameter:
14739 Set the interlacing filter coefficients. Accepts one of the following values:
14743 Simple filter coefficient set.
14745 More-complex filter coefficient set.
14747 Default value is @samp{complex}.
14750 Specify which frames to deinterlace. Accept one of the following values:
14754 Deinterlace all frames,
14756 Only deinterlace frames marked as interlaced.
14759 Default value is @samp{all}.
14763 Video waveform monitor.
14765 The waveform monitor plots color component intensity. By default luminance
14766 only. Each column of the waveform corresponds to a column of pixels in the
14769 It accepts the following options:
14773 Can be either @code{row}, or @code{column}. Default is @code{column}.
14774 In row mode, the graph on the left side represents color component value 0 and
14775 the right side represents value = 255. In column mode, the top side represents
14776 color component value = 0 and bottom side represents value = 255.
14779 Set intensity. Smaller values are useful to find out how many values of the same
14780 luminance are distributed across input rows/columns.
14781 Default value is @code{0.04}. Allowed range is [0, 1].
14784 Set mirroring mode. @code{0} means unmirrored, @code{1} means mirrored.
14785 In mirrored mode, higher values will be represented on the left
14786 side for @code{row} mode and at the top for @code{column} mode. Default is
14787 @code{1} (mirrored).
14791 It accepts the following values:
14794 Presents information identical to that in the @code{parade}, except
14795 that the graphs representing color components are superimposed directly
14798 This display mode makes it easier to spot relative differences or similarities
14799 in overlapping areas of the color components that are supposed to be identical,
14800 such as neutral whites, grays, or blacks.
14803 Display separate graph for the color components side by side in
14804 @code{row} mode or one below the other in @code{column} mode.
14807 Display separate graph for the color components side by side in
14808 @code{column} mode or one below the other in @code{row} mode.
14810 Using this display mode makes it easy to spot color casts in the highlights
14811 and shadows of an image, by comparing the contours of the top and the bottom
14812 graphs of each waveform. Since whites, grays, and blacks are characterized
14813 by exactly equal amounts of red, green, and blue, neutral areas of the picture
14814 should display three waveforms of roughly equal width/height. If not, the
14815 correction is easy to perform by making level adjustments the three waveforms.
14817 Default is @code{stack}.
14819 @item components, c
14820 Set which color components to display. Default is 1, which means only luminance
14821 or red color component if input is in RGB colorspace. If is set for example to
14822 7 it will display all 3 (if) available color components.
14827 No envelope, this is default.
14830 Instant envelope, minimum and maximum values presented in graph will be easily
14831 visible even with small @code{step} value.
14834 Hold minimum and maximum values presented in graph across time. This way you
14835 can still spot out of range values without constantly looking at waveforms.
14838 Peak and instant envelope combined together.
14844 No filtering, this is default.
14847 Luma and chroma combined together.
14850 Similar as above, but shows difference between blue and red chroma.
14853 Displays only chroma.
14856 Displays actual color value on waveform.
14859 Similar as above, but with luma showing frequency of chroma values.
14863 Set which graticule to display.
14867 Do not display graticule.
14870 Display green graticule showing legal broadcast ranges.
14874 Set graticule opacity.
14877 Set graticule flags.
14881 Draw numbers above lines. By default enabled.
14884 Draw dots instead of lines.
14888 Set scale used for displaying graticule.
14895 Default is digital.
14898 Set background opacity.
14901 @section weave, doubleweave
14903 The @code{weave} takes a field-based video input and join
14904 each two sequential fields into single frame, producing a new double
14905 height clip with half the frame rate and half the frame count.
14907 The @code{doubleweave} works same as @code{weave} but without
14908 halving frame rate and frame count.
14910 It accepts the following option:
14914 Set first field. Available values are:
14918 Set the frame as top-field-first.
14921 Set the frame as bottom-field-first.
14925 @subsection Examples
14929 Interlace video using @ref{select} and @ref{separatefields} filter:
14931 separatefields,select=eq(mod(n,4),0)+eq(mod(n,4),3),weave
14936 Apply the xBR high-quality magnification filter which is designed for pixel
14937 art. It follows a set of edge-detection rules, see
14938 @url{http://www.libretro.com/forums/viewtopic.php?f=6&t=134}.
14940 It accepts the following option:
14944 Set the scaling dimension: @code{2} for @code{2xBR}, @code{3} for
14945 @code{3xBR} and @code{4} for @code{4xBR}.
14946 Default is @code{3}.
14952 Deinterlace the input video ("yadif" means "yet another deinterlacing
14955 It accepts the following parameters:
14961 The interlacing mode to adopt. It accepts one of the following values:
14964 @item 0, send_frame
14965 Output one frame for each frame.
14966 @item 1, send_field
14967 Output one frame for each field.
14968 @item 2, send_frame_nospatial
14969 Like @code{send_frame}, but it skips the spatial interlacing check.
14970 @item 3, send_field_nospatial
14971 Like @code{send_field}, but it skips the spatial interlacing check.
14974 The default value is @code{send_frame}.
14977 The picture field parity assumed for the input interlaced video. It accepts one
14978 of the following values:
14982 Assume the top field is first.
14984 Assume the bottom field is first.
14986 Enable automatic detection of field parity.
14989 The default value is @code{auto}.
14990 If the interlacing is unknown or the decoder does not export this information,
14991 top field first will be assumed.
14994 Specify which frames to deinterlace. Accept one of the following
14999 Deinterlace all frames.
15000 @item 1, interlaced
15001 Only deinterlace frames marked as interlaced.
15004 The default value is @code{all}.
15009 Apply Zoom & Pan effect.
15011 This filter accepts the following options:
15015 Set the zoom expression. Default is 1.
15019 Set the x and y expression. Default is 0.
15022 Set the duration expression in number of frames.
15023 This sets for how many number of frames effect will last for
15024 single input image.
15027 Set the output image size, default is 'hd720'.
15030 Set the output frame rate, default is '25'.
15033 Each expression can contain the following constants:
15052 Output frame count.
15056 Last calculated 'x' and 'y' position from 'x' and 'y' expression
15057 for current input frame.
15061 'x' and 'y' of last output frame of previous input frame or 0 when there was
15062 not yet such frame (first input frame).
15065 Last calculated zoom from 'z' expression for current input frame.
15068 Last calculated zoom of last output frame of previous input frame.
15071 Number of output frames for current input frame. Calculated from 'd' expression
15072 for each input frame.
15075 number of output frames created for previous input frame
15078 Rational number: input width / input height
15081 sample aspect ratio
15084 display aspect ratio
15088 @subsection Examples
15092 Zoom-in up to 1.5 and pan at same time to some spot near center of picture:
15094 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='if(gte(zoom,1.5),x,x+1/a)':y='if(gte(zoom,1.5),y,y+1)':s=640x360
15098 Zoom-in up to 1.5 and pan always at center of picture:
15100 zoompan=z='min(zoom+0.0015,1.5)':d=700:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
15104 Same as above but without pausing:
15106 zoompan=z='min(max(zoom,pzoom)+0.0015,1.5)':d=1:x='iw/2-(iw/zoom/2)':y='ih/2-(ih/zoom/2)'
15111 Scale (resize) the input video, using the z.lib library:
15112 https://github.com/sekrit-twc/zimg.
15114 The zscale filter forces the output display aspect ratio to be the same
15115 as the input, by changing the output sample aspect ratio.
15117 If the input image format is different from the format requested by
15118 the next filter, the zscale filter will convert the input to the
15121 @subsection Options
15122 The filter accepts the following options.
15127 Set the output video dimension expression. Default value is the input
15130 If the @var{width} or @var{w} is 0, the input width is used for the output.
15131 If the @var{height} or @var{h} is 0, the input height is used for the output.
15133 If one of the values is -1, the zscale filter will use a value that
15134 maintains the aspect ratio of the input image, calculated from the
15135 other specified dimension. If both of them are -1, the input size is
15138 If one of the values is -n with n > 1, the zscale filter will also use a value
15139 that maintains the aspect ratio of the input image, calculated from the other
15140 specified dimension. After that it will, however, make sure that the calculated
15141 dimension is divisible by n and adjust the value if necessary.
15143 See below for the list of accepted constants for use in the dimension
15147 Set the video size. For the syntax of this option, check the
15148 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15151 Set the dither type.
15153 Possible values are:
15158 @item error_diffusion
15164 Set the resize filter type.
15166 Possible values are:
15176 Default is bilinear.
15179 Set the color range.
15181 Possible values are:
15188 Default is same as input.
15191 Set the color primaries.
15193 Possible values are:
15203 Default is same as input.
15206 Set the transfer characteristics.
15208 Possible values are:
15222 Default is same as input.
15225 Set the colorspace matrix.
15227 Possible value are:
15238 Default is same as input.
15241 Set the input color range.
15243 Possible values are:
15250 Default is same as input.
15252 @item primariesin, pin
15253 Set the input color primaries.
15255 Possible values are:
15265 Default is same as input.
15267 @item transferin, tin
15268 Set the input transfer characteristics.
15270 Possible values are:
15281 Default is same as input.
15283 @item matrixin, min
15284 Set the input colorspace matrix.
15286 Possible value are:
15298 Set the output chroma location.
15300 Possible values are:
15311 @item chromalin, cin
15312 Set the input chroma location.
15314 Possible values are:
15326 Set the nominal peak luminance.
15329 The values of the @option{w} and @option{h} options are expressions
15330 containing the following constants:
15335 The input width and height
15339 These are the same as @var{in_w} and @var{in_h}.
15343 The output (scaled) width and height
15347 These are the same as @var{out_w} and @var{out_h}
15350 The same as @var{iw} / @var{ih}
15353 input sample aspect ratio
15356 The input display aspect ratio. Calculated from @code{(iw / ih) * sar}.
15360 horizontal and vertical input chroma subsample values. For example for the
15361 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15365 horizontal and vertical output chroma subsample values. For example for the
15366 pixel format "yuv422p" @var{hsub} is 2 and @var{vsub} is 1.
15372 @c man end VIDEO FILTERS
15374 @chapter Video Sources
15375 @c man begin VIDEO SOURCES
15377 Below is a description of the currently available video sources.
15381 Buffer video frames, and make them available to the filter chain.
15383 This source is mainly intended for a programmatic use, in particular
15384 through the interface defined in @file{libavfilter/vsrc_buffer.h}.
15386 It accepts the following parameters:
15391 Specify the size (width and height) of the buffered video frames. For the
15392 syntax of this option, check the
15393 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15396 The input video width.
15399 The input video height.
15402 A string representing the pixel format of the buffered video frames.
15403 It may be a number corresponding to a pixel format, or a pixel format
15407 Specify the timebase assumed by the timestamps of the buffered frames.
15410 Specify the frame rate expected for the video stream.
15412 @item pixel_aspect, sar
15413 The sample (pixel) aspect ratio of the input video.
15416 Specify the optional parameters to be used for the scale filter which
15417 is automatically inserted when an input change is detected in the
15418 input size or format.
15420 @item hw_frames_ctx
15421 When using a hardware pixel format, this should be a reference to an
15422 AVHWFramesContext describing input frames.
15427 buffer=width=320:height=240:pix_fmt=yuv410p:time_base=1/24:sar=1
15430 will instruct the source to accept video frames with size 320x240 and
15431 with format "yuv410p", assuming 1/24 as the timestamps timebase and
15432 square pixels (1:1 sample aspect ratio).
15433 Since the pixel format with name "yuv410p" corresponds to the number 6
15434 (check the enum AVPixelFormat definition in @file{libavutil/pixfmt.h}),
15435 this example corresponds to:
15437 buffer=size=320x240:pixfmt=6:time_base=1/24:pixel_aspect=1/1
15440 Alternatively, the options can be specified as a flat string, but this
15441 syntax is deprecated:
15443 @var{width}:@var{height}:@var{pix_fmt}:@var{time_base.num}:@var{time_base.den}:@var{pixel_aspect.num}:@var{pixel_aspect.den}[:@var{sws_param}]
15447 Create a pattern generated by an elementary cellular automaton.
15449 The initial state of the cellular automaton can be defined through the
15450 @option{filename} and @option{pattern} options. If such options are
15451 not specified an initial state is created randomly.
15453 At each new frame a new row in the video is filled with the result of
15454 the cellular automaton next generation. The behavior when the whole
15455 frame is filled is defined by the @option{scroll} option.
15457 This source accepts the following options:
15461 Read the initial cellular automaton state, i.e. the starting row, from
15462 the specified file.
15463 In the file, each non-whitespace character is considered an alive
15464 cell, a newline will terminate the row, and further characters in the
15465 file will be ignored.
15468 Read the initial cellular automaton state, i.e. the starting row, from
15469 the specified string.
15471 Each non-whitespace character in the string is considered an alive
15472 cell, a newline will terminate the row, and further characters in the
15473 string will be ignored.
15476 Set the video rate, that is the number of frames generated per second.
15479 @item random_fill_ratio, ratio
15480 Set the random fill ratio for the initial cellular automaton row. It
15481 is a floating point number value ranging from 0 to 1, defaults to
15484 This option is ignored when a file or a pattern is specified.
15486 @item random_seed, seed
15487 Set the seed for filling randomly the initial row, must be an integer
15488 included between 0 and UINT32_MAX. If not specified, or if explicitly
15489 set to -1, the filter will try to use a good random seed on a best
15493 Set the cellular automaton rule, it is a number ranging from 0 to 255.
15494 Default value is 110.
15497 Set the size of the output video. For the syntax of this option, check the
15498 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15500 If @option{filename} or @option{pattern} is specified, the size is set
15501 by default to the width of the specified initial state row, and the
15502 height is set to @var{width} * PHI.
15504 If @option{size} is set, it must contain the width of the specified
15505 pattern string, and the specified pattern will be centered in the
15508 If a filename or a pattern string is not specified, the size value
15509 defaults to "320x518" (used for a randomly generated initial state).
15512 If set to 1, scroll the output upward when all the rows in the output
15513 have been already filled. If set to 0, the new generated row will be
15514 written over the top row just after the bottom row is filled.
15517 @item start_full, full
15518 If set to 1, completely fill the output with generated rows before
15519 outputting the first frame.
15520 This is the default behavior, for disabling set the value to 0.
15523 If set to 1, stitch the left and right row edges together.
15524 This is the default behavior, for disabling set the value to 0.
15527 @subsection Examples
15531 Read the initial state from @file{pattern}, and specify an output of
15534 cellauto=f=pattern:s=200x400
15538 Generate a random initial row with a width of 200 cells, with a fill
15541 cellauto=ratio=2/3:s=200x200
15545 Create a pattern generated by rule 18 starting by a single alive cell
15546 centered on an initial row with width 100:
15548 cellauto=p=@@:s=100x400:full=0:rule=18
15552 Specify a more elaborated initial pattern:
15554 cellauto=p='@@@@ @@ @@@@':s=100x400:full=0:rule=18
15559 @anchor{coreimagesrc}
15560 @section coreimagesrc
15561 Video source generated on GPU using Apple's CoreImage API on OSX.
15563 This video source is a specialized version of the @ref{coreimage} video filter.
15564 Use a core image generator at the beginning of the applied filterchain to
15565 generate the content.
15567 The coreimagesrc video source accepts the following options:
15569 @item list_generators
15570 List all available generators along with all their respective options as well as
15571 possible minimum and maximum values along with the default values.
15573 list_generators=true
15577 Specify the size of the sourced video. For the syntax of this option, check the
15578 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15579 The default value is @code{320x240}.
15582 Specify the frame rate of the sourced video, as the number of frames
15583 generated per second. It has to be a string in the format
15584 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15585 number or a valid video frame rate abbreviation. The default value is
15589 Set the sample aspect ratio of the sourced video.
15592 Set the duration of the sourced video. See
15593 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15594 for the accepted syntax.
15596 If not specified, or the expressed duration is negative, the video is
15597 supposed to be generated forever.
15600 Additionally, all options of the @ref{coreimage} video filter are accepted.
15601 A complete filterchain can be used for further processing of the
15602 generated input without CPU-HOST transfer. See @ref{coreimage} documentation
15603 and examples for details.
15605 @subsection Examples
15610 Use CIQRCodeGenerator to create a QR code for the FFmpeg homepage,
15611 given as complete and escaped command-line for Apple's standard bash shell:
15613 ffmpeg -f lavfi -i coreimagesrc=s=100x100:filter=CIQRCodeGenerator@@inputMessage=https\\\\\://FFmpeg.org/@@inputCorrectionLevel=H -frames:v 1 QRCode.png
15615 This example is equivalent to the QRCode example of @ref{coreimage} without the
15616 need for a nullsrc video source.
15620 @section mandelbrot
15622 Generate a Mandelbrot set fractal, and progressively zoom towards the
15623 point specified with @var{start_x} and @var{start_y}.
15625 This source accepts the following options:
15630 Set the terminal pts value. Default value is 400.
15633 Set the terminal scale value.
15634 Must be a floating point value. Default value is 0.3.
15637 Set the inner coloring mode, that is the algorithm used to draw the
15638 Mandelbrot fractal internal region.
15640 It shall assume one of the following values:
15645 Show time until convergence.
15647 Set color based on point closest to the origin of the iterations.
15652 Default value is @var{mincol}.
15655 Set the bailout value. Default value is 10.0.
15658 Set the maximum of iterations performed by the rendering
15659 algorithm. Default value is 7189.
15662 Set outer coloring mode.
15663 It shall assume one of following values:
15665 @item iteration_count
15666 Set iteration cound mode.
15667 @item normalized_iteration_count
15668 set normalized iteration count mode.
15670 Default value is @var{normalized_iteration_count}.
15673 Set frame rate, expressed as number of frames per second. Default
15677 Set frame size. For the syntax of this option, check the "Video
15678 size" section in the ffmpeg-utils manual. Default value is "640x480".
15681 Set the initial scale value. Default value is 3.0.
15684 Set the initial x position. Must be a floating point value between
15685 -100 and 100. Default value is -0.743643887037158704752191506114774.
15688 Set the initial y position. Must be a floating point value between
15689 -100 and 100. Default value is -0.131825904205311970493132056385139.
15694 Generate various test patterns, as generated by the MPlayer test filter.
15696 The size of the generated video is fixed, and is 256x256.
15697 This source is useful in particular for testing encoding features.
15699 This source accepts the following options:
15704 Specify the frame rate of the sourced video, as the number of frames
15705 generated per second. It has to be a string in the format
15706 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15707 number or a valid video frame rate abbreviation. The default value is
15711 Set the duration of the sourced video. See
15712 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15713 for the accepted syntax.
15715 If not specified, or the expressed duration is negative, the video is
15716 supposed to be generated forever.
15720 Set the number or the name of the test to perform. Supported tests are:
15736 Default value is "all", which will cycle through the list of all tests.
15741 mptestsrc=t=dc_luma
15744 will generate a "dc_luma" test pattern.
15746 @section frei0r_src
15748 Provide a frei0r source.
15750 To enable compilation of this filter you need to install the frei0r
15751 header and configure FFmpeg with @code{--enable-frei0r}.
15753 This source accepts the following parameters:
15758 The size of the video to generate. For the syntax of this option, check the
15759 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15762 The framerate of the generated video. It may be a string of the form
15763 @var{num}/@var{den} or a frame rate abbreviation.
15766 The name to the frei0r source to load. For more information regarding frei0r and
15767 how to set the parameters, read the @ref{frei0r} section in the video filters
15770 @item filter_params
15771 A '|'-separated list of parameters to pass to the frei0r source.
15775 For example, to generate a frei0r partik0l source with size 200x200
15776 and frame rate 10 which is overlaid on the overlay filter main input:
15778 frei0r_src=size=200x200:framerate=10:filter_name=partik0l:filter_params=1234 [overlay]; [in][overlay] overlay
15783 Generate a life pattern.
15785 This source is based on a generalization of John Conway's life game.
15787 The sourced input represents a life grid, each pixel represents a cell
15788 which can be in one of two possible states, alive or dead. Every cell
15789 interacts with its eight neighbours, which are the cells that are
15790 horizontally, vertically, or diagonally adjacent.
15792 At each interaction the grid evolves according to the adopted rule,
15793 which specifies the number of neighbor alive cells which will make a
15794 cell stay alive or born. The @option{rule} option allows one to specify
15797 This source accepts the following options:
15801 Set the file from which to read the initial grid state. In the file,
15802 each non-whitespace character is considered an alive cell, and newline
15803 is used to delimit the end of each row.
15805 If this option is not specified, the initial grid is generated
15809 Set the video rate, that is the number of frames generated per second.
15812 @item random_fill_ratio, ratio
15813 Set the random fill ratio for the initial random grid. It is a
15814 floating point number value ranging from 0 to 1, defaults to 1/PHI.
15815 It is ignored when a file is specified.
15817 @item random_seed, seed
15818 Set the seed for filling the initial random grid, must be an integer
15819 included between 0 and UINT32_MAX. If not specified, or if explicitly
15820 set to -1, the filter will try to use a good random seed on a best
15826 A rule can be specified with a code of the kind "S@var{NS}/B@var{NB}",
15827 where @var{NS} and @var{NB} are sequences of numbers in the range 0-8,
15828 @var{NS} specifies the number of alive neighbor cells which make a
15829 live cell stay alive, and @var{NB} the number of alive neighbor cells
15830 which make a dead cell to become alive (i.e. to "born").
15831 "s" and "b" can be used in place of "S" and "B", respectively.
15833 Alternatively a rule can be specified by an 18-bits integer. The 9
15834 high order bits are used to encode the next cell state if it is alive
15835 for each number of neighbor alive cells, the low order bits specify
15836 the rule for "borning" new cells. Higher order bits encode for an
15837 higher number of neighbor cells.
15838 For example the number 6153 = @code{(12<<9)+9} specifies a stay alive
15839 rule of 12 and a born rule of 9, which corresponds to "S23/B03".
15841 Default value is "S23/B3", which is the original Conway's game of life
15842 rule, and will keep a cell alive if it has 2 or 3 neighbor alive
15843 cells, and will born a new cell if there are three alive cells around
15847 Set the size of the output video. For the syntax of this option, check the
15848 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15850 If @option{filename} is specified, the size is set by default to the
15851 same size of the input file. If @option{size} is set, it must contain
15852 the size specified in the input file, and the initial grid defined in
15853 that file is centered in the larger resulting area.
15855 If a filename is not specified, the size value defaults to "320x240"
15856 (used for a randomly generated initial grid).
15859 If set to 1, stitch the left and right grid edges together, and the
15860 top and bottom edges also. Defaults to 1.
15863 Set cell mold speed. If set, a dead cell will go from @option{death_color} to
15864 @option{mold_color} with a step of @option{mold}. @option{mold} can have a
15865 value from 0 to 255.
15868 Set the color of living (or new born) cells.
15871 Set the color of dead cells. If @option{mold} is set, this is the first color
15872 used to represent a dead cell.
15875 Set mold color, for definitely dead and moldy cells.
15877 For the syntax of these 3 color options, check the "Color" section in the
15878 ffmpeg-utils manual.
15881 @subsection Examples
15885 Read a grid from @file{pattern}, and center it on a grid of size
15888 life=f=pattern:s=300x300
15892 Generate a random grid of size 200x200, with a fill ratio of 2/3:
15894 life=ratio=2/3:s=200x200
15898 Specify a custom rule for evolving a randomly generated grid:
15904 Full example with slow death effect (mold) using @command{ffplay}:
15906 ffplay -f lavfi life=s=300x200:mold=10:r=60:ratio=0.1:death_color=#C83232:life_color=#00ff00,scale=1200:800:flags=16
15913 @anchor{haldclutsrc}
15915 @anchor{rgbtestsrc}
15917 @anchor{smptehdbars}
15920 @anchor{yuvtestsrc}
15921 @section allrgb, allyuv, color, haldclutsrc, nullsrc, rgbtestsrc, smptebars, smptehdbars, testsrc, testsrc2, yuvtestsrc
15923 The @code{allrgb} source returns frames of size 4096x4096 of all rgb colors.
15925 The @code{allyuv} source returns frames of size 4096x4096 of all yuv colors.
15927 The @code{color} source provides an uniformly colored input.
15929 The @code{haldclutsrc} source provides an identity Hald CLUT. See also
15930 @ref{haldclut} filter.
15932 The @code{nullsrc} source returns unprocessed video frames. It is
15933 mainly useful to be employed in analysis / debugging tools, or as the
15934 source for filters which ignore the input data.
15936 The @code{rgbtestsrc} source generates an RGB test pattern useful for
15937 detecting RGB vs BGR issues. You should see a red, green and blue
15938 stripe from top to bottom.
15940 The @code{smptebars} source generates a color bars pattern, based on
15941 the SMPTE Engineering Guideline EG 1-1990.
15943 The @code{smptehdbars} source generates a color bars pattern, based on
15944 the SMPTE RP 219-2002.
15946 The @code{testsrc} source generates a test video pattern, showing a
15947 color pattern, a scrolling gradient and a timestamp. This is mainly
15948 intended for testing purposes.
15950 The @code{testsrc2} source is similar to testsrc, but supports more
15951 pixel formats instead of just @code{rgb24}. This allows using it as an
15952 input for other tests without requiring a format conversion.
15954 The @code{yuvtestsrc} source generates an YUV test pattern. You should
15955 see a y, cb and cr stripe from top to bottom.
15957 The sources accept the following parameters:
15962 Specify the color of the source, only available in the @code{color}
15963 source. For the syntax of this option, check the "Color" section in the
15964 ffmpeg-utils manual.
15967 Specify the level of the Hald CLUT, only available in the @code{haldclutsrc}
15968 source. A level of @code{N} generates a picture of @code{N*N*N} by @code{N*N*N}
15969 pixels to be used as identity matrix for 3D lookup tables. Each component is
15970 coded on a @code{1/(N*N)} scale.
15973 Specify the size of the sourced video. For the syntax of this option, check the
15974 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
15975 The default value is @code{320x240}.
15977 This option is not available with the @code{haldclutsrc} filter.
15980 Specify the frame rate of the sourced video, as the number of frames
15981 generated per second. It has to be a string in the format
15982 @var{frame_rate_num}/@var{frame_rate_den}, an integer number, a floating point
15983 number or a valid video frame rate abbreviation. The default value is
15987 Set the sample aspect ratio of the sourced video.
15990 Set the duration of the sourced video. See
15991 @ref{time duration syntax,,the Time duration section in the ffmpeg-utils(1) manual,ffmpeg-utils}
15992 for the accepted syntax.
15994 If not specified, or the expressed duration is negative, the video is
15995 supposed to be generated forever.
15998 Set the number of decimals to show in the timestamp, only available in the
15999 @code{testsrc} source.
16001 The displayed timestamp value will correspond to the original
16002 timestamp value multiplied by the power of 10 of the specified
16003 value. Default value is 0.
16006 For example the following:
16008 testsrc=duration=5.3:size=qcif:rate=10
16011 will generate a video with a duration of 5.3 seconds, with size
16012 176x144 and a frame rate of 10 frames per second.
16014 The following graph description will generate a red source
16015 with an opacity of 0.2, with size "qcif" and a frame rate of 10
16018 color=c=red@@0.2:s=qcif:r=10
16021 If the input content is to be ignored, @code{nullsrc} can be used. The
16022 following command generates noise in the luminance plane by employing
16023 the @code{geq} filter:
16025 nullsrc=s=256x256, geq=random(1)*255:128:128
16028 @subsection Commands
16030 The @code{color} source supports the following commands:
16034 Set the color of the created image. Accepts the same syntax of the
16035 corresponding @option{color} option.
16038 @c man end VIDEO SOURCES
16040 @chapter Video Sinks
16041 @c man begin VIDEO SINKS
16043 Below is a description of the currently available video sinks.
16045 @section buffersink
16047 Buffer video frames, and make them available to the end of the filter
16050 This sink is mainly intended for programmatic use, in particular
16051 through the interface defined in @file{libavfilter/buffersink.h}
16052 or the options system.
16054 It accepts a pointer to an AVBufferSinkContext structure, which
16055 defines the incoming buffers' formats, to be passed as the opaque
16056 parameter to @code{avfilter_init_filter} for initialization.
16060 Null video sink: do absolutely nothing with the input video. It is
16061 mainly useful as a template and for use in analysis / debugging
16064 @c man end VIDEO SINKS
16066 @chapter Multimedia Filters
16067 @c man begin MULTIMEDIA FILTERS
16069 Below is a description of the currently available multimedia filters.
16073 Convert input audio to a video output, displaying the audio bit scope.
16075 The filter accepts the following options:
16079 Set frame rate, expressed as number of frames per second. Default
16083 Specify the video size for the output. For the syntax of this option, check the
16084 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16085 Default value is @code{1024x256}.
16088 Specify list of colors separated by space or by '|' which will be used to
16089 draw channels. Unrecognized or missing colors will be replaced
16093 @section ahistogram
16095 Convert input audio to a video output, displaying the volume histogram.
16097 The filter accepts the following options:
16101 Specify how histogram is calculated.
16103 It accepts the following values:
16106 Use single histogram for all channels.
16108 Use separate histogram for each channel.
16110 Default is @code{single}.
16113 Set frame rate, expressed as number of frames per second. Default
16117 Specify the video size for the output. For the syntax of this option, check the
16118 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16119 Default value is @code{hd720}.
16124 It accepts the following values:
16135 reverse logarithmic
16137 Default is @code{log}.
16140 Set amplitude scale.
16142 It accepts the following values:
16149 Default is @code{log}.
16152 Set how much frames to accumulate in histogram.
16153 Defauls is 1. Setting this to -1 accumulates all frames.
16156 Set histogram ratio of window height.
16159 Set sonogram sliding.
16161 It accepts the following values:
16164 replace old rows with new ones.
16166 scroll from top to bottom.
16168 Default is @code{replace}.
16171 @section aphasemeter
16173 Convert input audio to a video output, displaying the audio phase.
16175 The filter accepts the following options:
16179 Set the output frame rate. Default value is @code{25}.
16182 Set the video size for the output. For the syntax of this option, check the
16183 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16184 Default value is @code{800x400}.
16189 Specify the red, green, blue contrast. Default values are @code{2},
16190 @code{7} and @code{1}.
16191 Allowed range is @code{[0, 255]}.
16194 Set color which will be used for drawing median phase. If color is
16195 @code{none} which is default, no median phase value will be drawn.
16198 Enable video output. Default is enabled.
16201 The filter also exports the frame metadata @code{lavfi.aphasemeter.phase} which
16202 represents mean phase of current audio frame. Value is in range @code{[-1, 1]}.
16203 The @code{-1} means left and right channels are completely out of phase and
16204 @code{1} means channels are in phase.
16206 @section avectorscope
16208 Convert input audio to a video output, representing the audio vector
16211 The filter is used to measure the difference between channels of stereo
16212 audio stream. A monoaural signal, consisting of identical left and right
16213 signal, results in straight vertical line. Any stereo separation is visible
16214 as a deviation from this line, creating a Lissajous figure.
16215 If the straight (or deviation from it) but horizontal line appears this
16216 indicates that the left and right channels are out of phase.
16218 The filter accepts the following options:
16222 Set the vectorscope mode.
16224 Available values are:
16227 Lissajous rotated by 45 degrees.
16230 Same as above but not rotated.
16233 Shape resembling half of circle.
16236 Default value is @samp{lissajous}.
16239 Set the video size for the output. For the syntax of this option, check the
16240 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16241 Default value is @code{400x400}.
16244 Set the output frame rate. Default value is @code{25}.
16250 Specify the red, green, blue and alpha contrast. Default values are @code{40},
16251 @code{160}, @code{80} and @code{255}.
16252 Allowed range is @code{[0, 255]}.
16258 Specify the red, green, blue and alpha fade. Default values are @code{15},
16259 @code{10}, @code{5} and @code{5}.
16260 Allowed range is @code{[0, 255]}.
16263 Set the zoom factor. Default value is @code{1}. Allowed range is @code{[1, 10]}.
16266 Set the vectorscope drawing mode.
16268 Available values are:
16271 Draw dot for each sample.
16274 Draw line between previous and current sample.
16277 Default value is @samp{dot}.
16280 Specify amplitude scale of audio samples.
16282 Available values are:
16299 @subsection Examples
16303 Complete example using @command{ffplay}:
16305 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
16306 [a] avectorscope=zoom=1.3:rc=2:gc=200:bc=10:rf=1:gf=8:bf=7 [out0]'
16310 @section bench, abench
16312 Benchmark part of a filtergraph.
16314 The filter accepts the following options:
16318 Start or stop a timer.
16320 Available values are:
16323 Get the current time, set it as frame metadata (using the key
16324 @code{lavfi.bench.start_time}), and forward the frame to the next filter.
16327 Get the current time and fetch the @code{lavfi.bench.start_time} metadata from
16328 the input frame metadata to get the time difference. Time difference, average,
16329 maximum and minimum time (respectively @code{t}, @code{avg}, @code{max} and
16330 @code{min}) are then printed. The timestamps are expressed in seconds.
16334 @subsection Examples
16338 Benchmark @ref{selectivecolor} filter:
16340 bench=start,selectivecolor=reds=-.2 .12 -.49,bench=stop
16346 Concatenate audio and video streams, joining them together one after the
16349 The filter works on segments of synchronized video and audio streams. All
16350 segments must have the same number of streams of each type, and that will
16351 also be the number of streams at output.
16353 The filter accepts the following options:
16358 Set the number of segments. Default is 2.
16361 Set the number of output video streams, that is also the number of video
16362 streams in each segment. Default is 1.
16365 Set the number of output audio streams, that is also the number of audio
16366 streams in each segment. Default is 0.
16369 Activate unsafe mode: do not fail if segments have a different format.
16373 The filter has @var{v}+@var{a} outputs: first @var{v} video outputs, then
16374 @var{a} audio outputs.
16376 There are @var{n}x(@var{v}+@var{a}) inputs: first the inputs for the first
16377 segment, in the same order as the outputs, then the inputs for the second
16380 Related streams do not always have exactly the same duration, for various
16381 reasons including codec frame size or sloppy authoring. For that reason,
16382 related synchronized streams (e.g. a video and its audio track) should be
16383 concatenated at once. The concat filter will use the duration of the longest
16384 stream in each segment (except the last one), and if necessary pad shorter
16385 audio streams with silence.
16387 For this filter to work correctly, all segments must start at timestamp 0.
16389 All corresponding streams must have the same parameters in all segments; the
16390 filtering system will automatically select a common pixel format for video
16391 streams, and a common sample format, sample rate and channel layout for
16392 audio streams, but other settings, such as resolution, must be converted
16393 explicitly by the user.
16395 Different frame rates are acceptable but will result in variable frame rate
16396 at output; be sure to configure the output file to handle it.
16398 @subsection Examples
16402 Concatenate an opening, an episode and an ending, all in bilingual version
16403 (video in stream 0, audio in streams 1 and 2):
16405 ffmpeg -i opening.mkv -i episode.mkv -i ending.mkv -filter_complex \
16406 '[0:0] [0:1] [0:2] [1:0] [1:1] [1:2] [2:0] [2:1] [2:2]
16407 concat=n=3:v=1:a=2 [v] [a1] [a2]' \
16408 -map '[v]' -map '[a1]' -map '[a2]' output.mkv
16412 Concatenate two parts, handling audio and video separately, using the
16413 (a)movie sources, and adjusting the resolution:
16415 movie=part1.mp4, scale=512:288 [v1] ; amovie=part1.mp4 [a1] ;
16416 movie=part2.mp4, scale=512:288 [v2] ; amovie=part2.mp4 [a2] ;
16417 [v1] [v2] concat [outv] ; [a1] [a2] concat=v=0:a=1 [outa]
16419 Note that a desync will happen at the stitch if the audio and video streams
16420 do not have exactly the same duration in the first file.
16424 @section drawgraph, adrawgraph
16426 Draw a graph using input video or audio metadata.
16428 It accepts the following parameters:
16432 Set 1st frame metadata key from which metadata values will be used to draw a graph.
16435 Set 1st foreground color expression.
16438 Set 2nd frame metadata key from which metadata values will be used to draw a graph.
16441 Set 2nd foreground color expression.
16444 Set 3rd frame metadata key from which metadata values will be used to draw a graph.
16447 Set 3rd foreground color expression.
16450 Set 4th frame metadata key from which metadata values will be used to draw a graph.
16453 Set 4th foreground color expression.
16456 Set minimal value of metadata value.
16459 Set maximal value of metadata value.
16462 Set graph background color. Default is white.
16467 Available values for mode is:
16474 Default is @code{line}.
16479 Available values for slide is:
16482 Draw new frame when right border is reached.
16485 Replace old columns with new ones.
16488 Scroll from right to left.
16491 Scroll from left to right.
16494 Draw single picture.
16497 Default is @code{frame}.
16500 Set size of graph video. For the syntax of this option, check the
16501 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16502 The default value is @code{900x256}.
16504 The foreground color expressions can use the following variables:
16507 Minimal value of metadata value.
16510 Maximal value of metadata value.
16513 Current metadata key value.
16516 The color is defined as 0xAABBGGRR.
16519 Example using metadata from @ref{signalstats} filter:
16521 signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255
16524 Example using metadata from @ref{ebur128} filter:
16526 ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5
16532 EBU R128 scanner filter. This filter takes an audio stream as input and outputs
16533 it unchanged. By default, it logs a message at a frequency of 10Hz with the
16534 Momentary loudness (identified by @code{M}), Short-term loudness (@code{S}),
16535 Integrated loudness (@code{I}) and Loudness Range (@code{LRA}).
16537 The filter also has a video output (see the @var{video} option) with a real
16538 time graph to observe the loudness evolution. The graphic contains the logged
16539 message mentioned above, so it is not printed anymore when this option is set,
16540 unless the verbose logging is set. The main graphing area contains the
16541 short-term loudness (3 seconds of analysis), and the gauge on the right is for
16542 the momentary loudness (400 milliseconds).
16544 More information about the Loudness Recommendation EBU R128 on
16545 @url{http://tech.ebu.ch/loudness}.
16547 The filter accepts the following options:
16552 Activate the video output. The audio stream is passed unchanged whether this
16553 option is set or no. The video stream will be the first output stream if
16554 activated. Default is @code{0}.
16557 Set the video size. This option is for video only. For the syntax of this
16559 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
16560 Default and minimum resolution is @code{640x480}.
16563 Set the EBU scale meter. Default is @code{9}. Common values are @code{9} and
16564 @code{18}, respectively for EBU scale meter +9 and EBU scale meter +18. Any
16565 other integer value between this range is allowed.
16568 Set metadata injection. If set to @code{1}, the audio input will be segmented
16569 into 100ms output frames, each of them containing various loudness information
16570 in metadata. All the metadata keys are prefixed with @code{lavfi.r128.}.
16572 Default is @code{0}.
16575 Force the frame logging level.
16577 Available values are:
16580 information logging level
16582 verbose logging level
16585 By default, the logging level is set to @var{info}. If the @option{video} or
16586 the @option{metadata} options are set, it switches to @var{verbose}.
16591 Available modes can be cumulated (the option is a @code{flag} type). Possible
16595 Disable any peak mode (default).
16597 Enable sample-peak mode.
16599 Simple peak mode looking for the higher sample value. It logs a message
16600 for sample-peak (identified by @code{SPK}).
16602 Enable true-peak mode.
16604 If enabled, the peak lookup is done on an over-sampled version of the input
16605 stream for better peak accuracy. It logs a message for true-peak.
16606 (identified by @code{TPK}) and true-peak per frame (identified by @code{FTPK}).
16607 This mode requires a build with @code{libswresample}.
16611 Treat mono input files as "dual mono". If a mono file is intended for playback
16612 on a stereo system, its EBU R128 measurement will be perceptually incorrect.
16613 If set to @code{true}, this option will compensate for this effect.
16614 Multi-channel input files are not affected by this option.
16617 Set a specific pan law to be used for the measurement of dual mono files.
16618 This parameter is optional, and has a default value of -3.01dB.
16621 @subsection Examples
16625 Real-time graph using @command{ffplay}, with a EBU scale meter +18:
16627 ffplay -f lavfi -i "amovie=input.mp3,ebur128=video=1:meter=18 [out0][out1]"
16631 Run an analysis with @command{ffmpeg}:
16633 ffmpeg -nostats -i input.mp3 -filter_complex ebur128 -f null -
16637 @section interleave, ainterleave
16639 Temporally interleave frames from several inputs.
16641 @code{interleave} works with video inputs, @code{ainterleave} with audio.
16643 These filters read frames from several inputs and send the oldest
16644 queued frame to the output.
16646 Input streams must have well defined, monotonically increasing frame
16649 In order to submit one frame to output, these filters need to enqueue
16650 at least one frame for each input, so they cannot work in case one
16651 input is not yet terminated and will not receive incoming frames.
16653 For example consider the case when one input is a @code{select} filter
16654 which always drops input frames. The @code{interleave} filter will keep
16655 reading from that input, but it will never be able to send new frames
16656 to output until the input sends an end-of-stream signal.
16658 Also, depending on inputs synchronization, the filters will drop
16659 frames in case one input receives more frames than the other ones, and
16660 the queue is already filled.
16662 These filters accept the following options:
16666 Set the number of different inputs, it is 2 by default.
16669 @subsection Examples
16673 Interleave frames belonging to different streams using @command{ffmpeg}:
16675 ffmpeg -i bambi.avi -i pr0n.mkv -filter_complex "[0:v][1:v] interleave" out.avi
16679 Add flickering blur effect:
16681 select='if(gt(random(0), 0.2), 1, 2)':n=2 [tmp], boxblur=2:2, [tmp] interleave
16685 @section metadata, ametadata
16687 Manipulate frame metadata.
16689 This filter accepts the following options:
16693 Set mode of operation of the filter.
16695 Can be one of the following:
16699 If both @code{value} and @code{key} is set, select frames
16700 which have such metadata. If only @code{key} is set, select
16701 every frame that has such key in metadata.
16704 Add new metadata @code{key} and @code{value}. If key is already available
16708 Modify value of already present key.
16711 If @code{value} is set, delete only keys that have such value.
16712 Otherwise, delete key. If @code{key} is not set, delete all metadata values in
16716 Print key and its value if metadata was found. If @code{key} is not set print all
16717 metadata values available in frame.
16721 Set key used with all modes. Must be set for all modes except @code{print} and @code{delete}.
16724 Set metadata value which will be used. This option is mandatory for
16725 @code{modify} and @code{add} mode.
16728 Which function to use when comparing metadata value and @code{value}.
16730 Can be one of following:
16734 Values are interpreted as strings, returns true if metadata value is same as @code{value}.
16737 Values are interpreted as strings, returns true if metadata value starts with
16738 the @code{value} option string.
16741 Values are interpreted as floats, returns true if metadata value is less than @code{value}.
16744 Values are interpreted as floats, returns true if @code{value} is equal with metadata value.
16747 Values are interpreted as floats, returns true if metadata value is greater than @code{value}.
16750 Values are interpreted as floats, returns true if expression from option @code{expr}
16755 Set expression which is used when @code{function} is set to @code{expr}.
16756 The expression is evaluated through the eval API and can contain the following
16761 Float representation of @code{value} from metadata key.
16764 Float representation of @code{value} as supplied by user in @code{value} option.
16768 If specified in @code{print} mode, output is written to the named file. Instead of
16769 plain filename any writable url can be specified. Filename ``-'' is a shorthand
16770 for standard output. If @code{file} option is not set, output is written to the log
16771 with AV_LOG_INFO loglevel.
16775 @subsection Examples
16779 Print all metadata values for frames with key @code{lavfi.singnalstats.YDIF} with values
16782 signalstats,metadata=print:key=lavfi.signalstats.YDIF:value=0:function=expr:expr='between(VALUE1,0,1)'
16785 Print silencedetect output to file @file{metadata.txt}.
16787 silencedetect,ametadata=mode=print:file=metadata.txt
16790 Direct all metadata to a pipe with file descriptor 4.
16792 metadata=mode=print:file='pipe\:4'
16796 @section perms, aperms
16798 Set read/write permissions for the output frames.
16800 These filters are mainly aimed at developers to test direct path in the
16801 following filter in the filtergraph.
16803 The filters accept the following options:
16807 Select the permissions mode.
16809 It accepts the following values:
16812 Do nothing. This is the default.
16814 Set all the output frames read-only.
16816 Set all the output frames directly writable.
16818 Make the frame read-only if writable, and writable if read-only.
16820 Set each output frame read-only or writable randomly.
16824 Set the seed for the @var{random} mode, must be an integer included between
16825 @code{0} and @code{UINT32_MAX}. If not specified, or if explicitly set to
16826 @code{-1}, the filter will try to use a good random seed on a best effort
16830 Note: in case of auto-inserted filter between the permission filter and the
16831 following one, the permission might not be received as expected in that
16832 following filter. Inserting a @ref{format} or @ref{aformat} filter before the
16833 perms/aperms filter can avoid this problem.
16835 @section realtime, arealtime
16837 Slow down filtering to match real time approximatively.
16839 These filters will pause the filtering for a variable amount of time to
16840 match the output rate with the input timestamps.
16841 They are similar to the @option{re} option to @code{ffmpeg}.
16843 They accept the following options:
16847 Time limit for the pauses. Any pause longer than that will be considered
16848 a timestamp discontinuity and reset the timer. Default is 2 seconds.
16852 @section select, aselect
16854 Select frames to pass in output.
16856 This filter accepts the following options:
16861 Set expression, which is evaluated for each input frame.
16863 If the expression is evaluated to zero, the frame is discarded.
16865 If the evaluation result is negative or NaN, the frame is sent to the
16866 first output; otherwise it is sent to the output with index
16867 @code{ceil(val)-1}, assuming that the input index starts from 0.
16869 For example a value of @code{1.2} corresponds to the output with index
16870 @code{ceil(1.2)-1 = 2-1 = 1}, that is the second output.
16873 Set the number of outputs. The output to which to send the selected
16874 frame is based on the result of the evaluation. Default value is 1.
16877 The expression can contain the following constants:
16881 The (sequential) number of the filtered frame, starting from 0.
16884 The (sequential) number of the selected frame, starting from 0.
16886 @item prev_selected_n
16887 The sequential number of the last selected frame. It's NAN if undefined.
16890 The timebase of the input timestamps.
16893 The PTS (Presentation TimeStamp) of the filtered video frame,
16894 expressed in @var{TB} units. It's NAN if undefined.
16897 The PTS of the filtered video frame,
16898 expressed in seconds. It's NAN if undefined.
16901 The PTS of the previously filtered video frame. It's NAN if undefined.
16903 @item prev_selected_pts
16904 The PTS of the last previously filtered video frame. It's NAN if undefined.
16906 @item prev_selected_t
16907 The PTS of the last previously selected video frame. It's NAN if undefined.
16910 The PTS of the first video frame in the video. It's NAN if undefined.
16913 The time of the first video frame in the video. It's NAN if undefined.
16915 @item pict_type @emph{(video only)}
16916 The type of the filtered frame. It can assume one of the following
16928 @item interlace_type @emph{(video only)}
16929 The frame interlace type. It can assume one of the following values:
16932 The frame is progressive (not interlaced).
16934 The frame is top-field-first.
16936 The frame is bottom-field-first.
16939 @item consumed_sample_n @emph{(audio only)}
16940 the number of selected samples before the current frame
16942 @item samples_n @emph{(audio only)}
16943 the number of samples in the current frame
16945 @item sample_rate @emph{(audio only)}
16946 the input sample rate
16949 This is 1 if the filtered frame is a key-frame, 0 otherwise.
16952 the position in the file of the filtered frame, -1 if the information
16953 is not available (e.g. for synthetic video)
16955 @item scene @emph{(video only)}
16956 value between 0 and 1 to indicate a new scene; a low value reflects a low
16957 probability for the current frame to introduce a new scene, while a higher
16958 value means the current frame is more likely to be one (see the example below)
16960 @item concatdec_select
16961 The concat demuxer can select only part of a concat input file by setting an
16962 inpoint and an outpoint, but the output packets may not be entirely contained
16963 in the selected interval. By using this variable, it is possible to skip frames
16964 generated by the concat demuxer which are not exactly contained in the selected
16967 This works by comparing the frame pts against the @var{lavf.concat.start_time}
16968 and the @var{lavf.concat.duration} packet metadata values which are also
16969 present in the decoded frames.
16971 The @var{concatdec_select} variable is -1 if the frame pts is at least
16972 start_time and either the duration metadata is missing or the frame pts is less
16973 than start_time + duration, 0 otherwise, and NaN if the start_time metadata is
16976 That basically means that an input frame is selected if its pts is within the
16977 interval set by the concat demuxer.
16981 The default value of the select expression is "1".
16983 @subsection Examples
16987 Select all frames in input:
16992 The example above is the same as:
17004 Select only I-frames:
17006 select='eq(pict_type\,I)'
17010 Select one frame every 100:
17012 select='not(mod(n\,100))'
17016 Select only frames contained in the 10-20 time interval:
17018 select=between(t\,10\,20)
17022 Select only I-frames contained in the 10-20 time interval:
17024 select=between(t\,10\,20)*eq(pict_type\,I)
17028 Select frames with a minimum distance of 10 seconds:
17030 select='isnan(prev_selected_t)+gte(t-prev_selected_t\,10)'
17034 Use aselect to select only audio frames with samples number > 100:
17036 aselect='gt(samples_n\,100)'
17040 Create a mosaic of the first scenes:
17042 ffmpeg -i video.avi -vf select='gt(scene\,0.4)',scale=160:120,tile -frames:v 1 preview.png
17045 Comparing @var{scene} against a value between 0.3 and 0.5 is generally a sane
17049 Send even and odd frames to separate outputs, and compose them:
17051 select=n=2:e='mod(n, 2)+1' [odd][even]; [odd] pad=h=2*ih [tmp]; [tmp][even] overlay=y=h
17055 Select useful frames from an ffconcat file which is using inpoints and
17056 outpoints but where the source files are not intra frame only.
17058 ffmpeg -copyts -vsync 0 -segment_time_metadata 1 -i input.ffconcat -vf select=concatdec_select -af aselect=concatdec_select output.avi
17062 @section sendcmd, asendcmd
17064 Send commands to filters in the filtergraph.
17066 These filters read commands to be sent to other filters in the
17069 @code{sendcmd} must be inserted between two video filters,
17070 @code{asendcmd} must be inserted between two audio filters, but apart
17071 from that they act the same way.
17073 The specification of commands can be provided in the filter arguments
17074 with the @var{commands} option, or in a file specified by the
17075 @var{filename} option.
17077 These filters accept the following options:
17080 Set the commands to be read and sent to the other filters.
17082 Set the filename of the commands to be read and sent to the other
17086 @subsection Commands syntax
17088 A commands description consists of a sequence of interval
17089 specifications, comprising a list of commands to be executed when a
17090 particular event related to that interval occurs. The occurring event
17091 is typically the current frame time entering or leaving a given time
17094 An interval is specified by the following syntax:
17096 @var{START}[-@var{END}] @var{COMMANDS};
17099 The time interval is specified by the @var{START} and @var{END} times.
17100 @var{END} is optional and defaults to the maximum time.
17102 The current frame time is considered within the specified interval if
17103 it is included in the interval [@var{START}, @var{END}), that is when
17104 the time is greater or equal to @var{START} and is lesser than
17107 @var{COMMANDS} consists of a sequence of one or more command
17108 specifications, separated by ",", relating to that interval. The
17109 syntax of a command specification is given by:
17111 [@var{FLAGS}] @var{TARGET} @var{COMMAND} @var{ARG}
17114 @var{FLAGS} is optional and specifies the type of events relating to
17115 the time interval which enable sending the specified command, and must
17116 be a non-null sequence of identifier flags separated by "+" or "|" and
17117 enclosed between "[" and "]".
17119 The following flags are recognized:
17122 The command is sent when the current frame timestamp enters the
17123 specified interval. In other words, the command is sent when the
17124 previous frame timestamp was not in the given interval, and the
17128 The command is sent when the current frame timestamp leaves the
17129 specified interval. In other words, the command is sent when the
17130 previous frame timestamp was in the given interval, and the
17134 If @var{FLAGS} is not specified, a default value of @code{[enter]} is
17137 @var{TARGET} specifies the target of the command, usually the name of
17138 the filter class or a specific filter instance name.
17140 @var{COMMAND} specifies the name of the command for the target filter.
17142 @var{ARG} is optional and specifies the optional list of argument for
17143 the given @var{COMMAND}.
17145 Between one interval specification and another, whitespaces, or
17146 sequences of characters starting with @code{#} until the end of line,
17147 are ignored and can be used to annotate comments.
17149 A simplified BNF description of the commands specification syntax
17152 @var{COMMAND_FLAG} ::= "enter" | "leave"
17153 @var{COMMAND_FLAGS} ::= @var{COMMAND_FLAG} [(+|"|")@var{COMMAND_FLAG}]
17154 @var{COMMAND} ::= ["[" @var{COMMAND_FLAGS} "]"] @var{TARGET} @var{COMMAND} [@var{ARG}]
17155 @var{COMMANDS} ::= @var{COMMAND} [,@var{COMMANDS}]
17156 @var{INTERVAL} ::= @var{START}[-@var{END}] @var{COMMANDS}
17157 @var{INTERVALS} ::= @var{INTERVAL}[;@var{INTERVALS}]
17160 @subsection Examples
17164 Specify audio tempo change at second 4:
17166 asendcmd=c='4.0 atempo tempo 1.5',atempo
17170 Specify a list of drawtext and hue commands in a file.
17172 # show text in the interval 5-10
17173 5.0-10.0 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=hello world',
17174 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=';
17176 # desaturate the image in the interval 15-20
17177 15.0-20.0 [enter] hue s 0,
17178 [enter] drawtext reinit 'fontfile=FreeSerif.ttf:text=nocolor',
17180 [leave] drawtext reinit 'fontfile=FreeSerif.ttf:text=color';
17182 # apply an exponential saturation fade-out effect, starting from time 25
17183 25 [enter] hue s exp(25-t)
17186 A filtergraph allowing to read and process the above command list
17187 stored in a file @file{test.cmd}, can be specified with:
17189 sendcmd=f=test.cmd,drawtext=fontfile=FreeSerif.ttf:text='',hue
17194 @section setpts, asetpts
17196 Change the PTS (presentation timestamp) of the input frames.
17198 @code{setpts} works on video frames, @code{asetpts} on audio frames.
17200 This filter accepts the following options:
17205 The expression which is evaluated for each frame to construct its timestamp.
17209 The expression is evaluated through the eval API and can contain the following
17214 frame rate, only defined for constant frame-rate video
17217 The presentation timestamp in input
17220 The count of the input frame for video or the number of consumed samples,
17221 not including the current frame for audio, starting from 0.
17223 @item NB_CONSUMED_SAMPLES
17224 The number of consumed samples, not including the current frame (only
17227 @item NB_SAMPLES, S
17228 The number of samples in the current frame (only audio)
17230 @item SAMPLE_RATE, SR
17231 The audio sample rate.
17234 The PTS of the first frame.
17237 the time in seconds of the first frame
17240 State whether the current frame is interlaced.
17243 the time in seconds of the current frame
17246 original position in the file of the frame, or undefined if undefined
17247 for the current frame
17250 The previous input PTS.
17253 previous input time in seconds
17256 The previous output PTS.
17259 previous output time in seconds
17262 The wallclock (RTC) time in microseconds. This is deprecated, use time(0)
17266 The wallclock (RTC) time at the start of the movie in microseconds.
17269 The timebase of the input timestamps.
17273 @subsection Examples
17277 Start counting PTS from zero
17279 setpts=PTS-STARTPTS
17283 Apply fast motion effect:
17289 Apply slow motion effect:
17295 Set fixed rate of 25 frames per second:
17301 Set fixed rate 25 fps with some jitter:
17303 setpts='1/(25*TB) * (N + 0.05 * sin(N*2*PI/25))'
17307 Apply an offset of 10 seconds to the input PTS:
17313 Generate timestamps from a "live source" and rebase onto the current timebase:
17315 setpts='(RTCTIME - RTCSTART) / (TB * 1000000)'
17319 Generate timestamps by counting samples:
17326 @section settb, asettb
17328 Set the timebase to use for the output frames timestamps.
17329 It is mainly useful for testing timebase configuration.
17331 It accepts the following parameters:
17336 The expression which is evaluated into the output timebase.
17340 The value for @option{tb} is an arithmetic expression representing a
17341 rational. The expression can contain the constants "AVTB" (the default
17342 timebase), "intb" (the input timebase) and "sr" (the sample rate,
17343 audio only). Default value is "intb".
17345 @subsection Examples
17349 Set the timebase to 1/25:
17355 Set the timebase to 1/10:
17361 Set the timebase to 1001/1000:
17367 Set the timebase to 2*intb:
17373 Set the default timebase value:
17380 Convert input audio to a video output representing frequency spectrum
17381 logarithmically using Brown-Puckette constant Q transform algorithm with
17382 direct frequency domain coefficient calculation (but the transform itself
17383 is not really constant Q, instead the Q factor is actually variable/clamped),
17384 with musical tone scale, from E0 to D#10.
17386 The filter accepts the following options:
17390 Specify the video size for the output. It must be even. For the syntax of this option,
17391 check the @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17392 Default value is @code{1920x1080}.
17395 Set the output frame rate. Default value is @code{25}.
17398 Set the bargraph height. It must be even. Default value is @code{-1} which
17399 computes the bargraph height automatically.
17402 Set the axis height. It must be even. Default value is @code{-1} which computes
17403 the axis height automatically.
17406 Set the sonogram height. It must be even. Default value is @code{-1} which
17407 computes the sonogram height automatically.
17410 Set the fullhd resolution. This option is deprecated, use @var{size}, @var{s}
17411 instead. Default value is @code{1}.
17413 @item sono_v, volume
17414 Specify the sonogram volume expression. It can contain variables:
17417 the @var{bar_v} evaluated expression
17418 @item frequency, freq, f
17419 the frequency where it is evaluated
17420 @item timeclamp, tc
17421 the value of @var{timeclamp} option
17425 @item a_weighting(f)
17426 A-weighting of equal loudness
17427 @item b_weighting(f)
17428 B-weighting of equal loudness
17429 @item c_weighting(f)
17430 C-weighting of equal loudness.
17432 Default value is @code{16}.
17434 @item bar_v, volume2
17435 Specify the bargraph volume expression. It can contain variables:
17438 the @var{sono_v} evaluated expression
17439 @item frequency, freq, f
17440 the frequency where it is evaluated
17441 @item timeclamp, tc
17442 the value of @var{timeclamp} option
17446 @item a_weighting(f)
17447 A-weighting of equal loudness
17448 @item b_weighting(f)
17449 B-weighting of equal loudness
17450 @item c_weighting(f)
17451 C-weighting of equal loudness.
17453 Default value is @code{sono_v}.
17455 @item sono_g, gamma
17456 Specify the sonogram gamma. Lower gamma makes the spectrum more contrast,
17457 higher gamma makes the spectrum having more range. Default value is @code{3}.
17458 Acceptable range is @code{[1, 7]}.
17460 @item bar_g, gamma2
17461 Specify the bargraph gamma. Default value is @code{1}. Acceptable range is
17465 Specify the bargraph transparency level. Lower value makes the bargraph sharper.
17466 Default value is @code{1}. Acceptable range is @code{[0, 1]}.
17468 @item timeclamp, tc
17469 Specify the transform timeclamp. At low frequency, there is trade-off between
17470 accuracy in time domain and frequency domain. If timeclamp is lower,
17471 event in time domain is represented more accurately (such as fast bass drum),
17472 otherwise event in frequency domain is represented more accurately
17473 (such as bass guitar). Acceptable range is @code{[0.002, 1]}. Default value is @code{0.17}.
17476 Set attack time in seconds. The default is @code{0} (disabled). Otherwise, it
17477 limits future samples by applying asymmetric windowing in time domain, useful
17478 when low latency is required. Accepted range is @code{[0, 1]}.
17481 Specify the transform base frequency. Default value is @code{20.01523126408007475},
17482 which is frequency 50 cents below E0. Acceptable range is @code{[10, 100000]}.
17485 Specify the transform end frequency. Default value is @code{20495.59681441799654},
17486 which is frequency 50 cents above D#10. Acceptable range is @code{[10, 100000]}.
17489 This option is deprecated and ignored.
17492 Specify the transform length in time domain. Use this option to control accuracy
17493 trade-off between time domain and frequency domain at every frequency sample.
17494 It can contain variables:
17496 @item frequency, freq, f
17497 the frequency where it is evaluated
17498 @item timeclamp, tc
17499 the value of @var{timeclamp} option.
17501 Default value is @code{384*tc/(384+tc*f)}.
17504 Specify the transform count for every video frame. Default value is @code{6}.
17505 Acceptable range is @code{[1, 30]}.
17508 Specify the transform count for every single pixel. Default value is @code{0},
17509 which makes it computed automatically. Acceptable range is @code{[0, 10]}.
17512 Specify font file for use with freetype to draw the axis. If not specified,
17513 use embedded font. Note that drawing with font file or embedded font is not
17514 implemented with custom @var{basefreq} and @var{endfreq}, use @var{axisfile}
17518 Specify fontconfig pattern. This has lower priority than @var{fontfile}.
17519 The : in the pattern may be replaced by | to avoid unnecessary escaping.
17522 Specify font color expression. This is arithmetic expression that should return
17523 integer value 0xRRGGBB. It can contain variables:
17525 @item frequency, freq, f
17526 the frequency where it is evaluated
17527 @item timeclamp, tc
17528 the value of @var{timeclamp} option
17533 midi number of frequency f, some midi numbers: E0(16), C1(24), C2(36), A4(69)
17534 @item r(x), g(x), b(x)
17535 red, green, and blue value of intensity x.
17537 Default value is @code{st(0, (midi(f)-59.5)/12);
17538 st(1, if(between(ld(0),0,1), 0.5-0.5*cos(2*PI*ld(0)), 0));
17539 r(1-ld(1)) + b(ld(1))}.
17542 Specify image file to draw the axis. This option override @var{fontfile} and
17543 @var{fontcolor} option.
17546 Enable/disable drawing text to the axis. If it is set to @code{0}, drawing to
17547 the axis is disabled, ignoring @var{fontfile} and @var{axisfile} option.
17548 Default value is @code{1}.
17551 Set colorspace. The accepted values are:
17554 Unspecified (default)
17563 BT.470BG or BT.601-6 625
17566 SMPTE-170M or BT.601-6 525
17572 BT.2020 with non-constant luminance
17577 Set spectrogram color scheme. This is list of floating point values with format
17578 @code{left_r|left_g|left_b|right_r|right_g|right_b}.
17579 The default is @code{1|0.5|0|0|0.5|1}.
17583 @subsection Examples
17587 Playing audio while showing the spectrum:
17589 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt [out0]'
17593 Same as above, but with frame rate 30 fps:
17595 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=fps=30:count=5 [out0]'
17599 Playing at 1280x720:
17601 ffplay -f lavfi 'amovie=a.mp3, asplit [a][out1]; [a] showcqt=s=1280x720:count=4 [out0]'
17605 Disable sonogram display:
17611 A1 and its harmonics: A1, A2, (near)E3, A3:
17613 ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
17614 asplit[a][out1]; [a] showcqt [out0]'
17618 Same as above, but with more accuracy in frequency domain:
17620 ffplay -f lavfi 'aevalsrc=0.1*sin(2*PI*55*t)+0.1*sin(4*PI*55*t)+0.1*sin(6*PI*55*t)+0.1*sin(8*PI*55*t),
17621 asplit[a][out1]; [a] showcqt=timeclamp=0.5 [out0]'
17627 bar_v=10:sono_v=bar_v*a_weighting(f)
17631 Custom gamma, now spectrum is linear to the amplitude.
17637 Custom tlength equation:
17639 tc=0.33:tlength='st(0,0.17); 384*tc / (384 / ld(0) + tc*f /(1-ld(0))) + 384*tc / (tc*f / ld(0) + 384 /(1-ld(0)))'
17643 Custom fontcolor and fontfile, C-note is colored green, others are colored blue:
17645 fontcolor='if(mod(floor(midi(f)+0.5),12), 0x0000FF, g(1))':fontfile=myfont.ttf
17649 Custom font using fontconfig:
17651 font='Courier New,Monospace,mono|bold'
17655 Custom frequency range with custom axis using image file:
17657 axisfile=myaxis.png:basefreq=40:endfreq=10000
17663 Convert input audio to video output representing the audio power spectrum.
17664 Audio amplitude is on Y-axis while frequency is on X-axis.
17666 The filter accepts the following options:
17670 Specify size of video. For the syntax of this option, check the
17671 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17672 Default is @code{1024x512}.
17676 This set how each frequency bin will be represented.
17678 It accepts the following values:
17684 Default is @code{bar}.
17687 Set amplitude scale.
17689 It accepts the following values:
17703 Default is @code{log}.
17706 Set frequency scale.
17708 It accepts the following values:
17717 Reverse logarithmic scale.
17719 Default is @code{lin}.
17724 It accepts the following values:
17740 Default is @code{w2048}
17743 Set windowing function.
17745 It accepts the following values:
17767 Default is @code{hanning}.
17770 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
17771 which means optimal overlap for selected window function will be picked.
17774 Set time averaging. Setting this to 0 will display current maximal peaks.
17775 Default is @code{1}, which means time averaging is disabled.
17778 Specify list of colors separated by space or by '|' which will be used to
17779 draw channel frequencies. Unrecognized or missing colors will be replaced
17783 Set channel display mode.
17785 It accepts the following values:
17790 Default is @code{combined}.
17793 Set minimum amplitude used in @code{log} amplitude scaler.
17797 @anchor{showspectrum}
17798 @section showspectrum
17800 Convert input audio to a video output, representing the audio frequency
17803 The filter accepts the following options:
17807 Specify the video size for the output. For the syntax of this option, check the
17808 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17809 Default value is @code{640x512}.
17812 Specify how the spectrum should slide along the window.
17814 It accepts the following values:
17817 the samples start again on the left when they reach the right
17819 the samples scroll from right to left
17821 frames are only produced when the samples reach the right
17823 the samples scroll from left to right
17826 Default value is @code{replace}.
17829 Specify display mode.
17831 It accepts the following values:
17834 all channels are displayed in the same row
17836 all channels are displayed in separate rows
17839 Default value is @samp{combined}.
17842 Specify display color mode.
17844 It accepts the following values:
17847 each channel is displayed in a separate color
17849 each channel is displayed using the same color scheme
17851 each channel is displayed using the rainbow color scheme
17853 each channel is displayed using the moreland color scheme
17855 each channel is displayed using the nebulae color scheme
17857 each channel is displayed using the fire color scheme
17859 each channel is displayed using the fiery color scheme
17861 each channel is displayed using the fruit color scheme
17863 each channel is displayed using the cool color scheme
17866 Default value is @samp{channel}.
17869 Specify scale used for calculating intensity color values.
17871 It accepts the following values:
17876 square root, default
17887 Default value is @samp{sqrt}.
17890 Set saturation modifier for displayed colors. Negative values provide
17891 alternative color scheme. @code{0} is no saturation at all.
17892 Saturation must be in [-10.0, 10.0] range.
17893 Default value is @code{1}.
17896 Set window function.
17898 It accepts the following values:
17922 Default value is @code{hann}.
17925 Set orientation of time vs frequency axis. Can be @code{vertical} or
17926 @code{horizontal}. Default is @code{vertical}.
17929 Set ratio of overlap window. Default value is @code{0}.
17930 When value is @code{1} overlap is set to recommended size for specific
17931 window function currently used.
17934 Set scale gain for calculating intensity color values.
17935 Default value is @code{1}.
17938 Set which data to display. Can be @code{magnitude}, default or @code{phase}.
17941 Set color rotation, must be in [-1.0, 1.0] range.
17942 Default value is @code{0}.
17945 The usage is very similar to the showwaves filter; see the examples in that
17948 @subsection Examples
17952 Large window with logarithmic color scaling:
17954 showspectrum=s=1280x480:scale=log
17958 Complete example for a colored and sliding spectrum per channel using @command{ffplay}:
17960 ffplay -f lavfi 'amovie=input.mp3, asplit [a][out1];
17961 [a] showspectrum=mode=separate:color=intensity:slide=1:scale=cbrt [out0]'
17965 @section showspectrumpic
17967 Convert input audio to a single video frame, representing the audio frequency
17970 The filter accepts the following options:
17974 Specify the video size for the output. For the syntax of this option, check the
17975 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
17976 Default value is @code{4096x2048}.
17979 Specify display mode.
17981 It accepts the following values:
17984 all channels are displayed in the same row
17986 all channels are displayed in separate rows
17988 Default value is @samp{combined}.
17991 Specify display color mode.
17993 It accepts the following values:
17996 each channel is displayed in a separate color
17998 each channel is displayed using the same color scheme
18000 each channel is displayed using the rainbow color scheme
18002 each channel is displayed using the moreland color scheme
18004 each channel is displayed using the nebulae color scheme
18006 each channel is displayed using the fire color scheme
18008 each channel is displayed using the fiery color scheme
18010 each channel is displayed using the fruit color scheme
18012 each channel is displayed using the cool color scheme
18014 Default value is @samp{intensity}.
18017 Specify scale used for calculating intensity color values.
18019 It accepts the following values:
18024 square root, default
18034 Default value is @samp{log}.
18037 Set saturation modifier for displayed colors. Negative values provide
18038 alternative color scheme. @code{0} is no saturation at all.
18039 Saturation must be in [-10.0, 10.0] range.
18040 Default value is @code{1}.
18043 Set window function.
18045 It accepts the following values:
18068 Default value is @code{hann}.
18071 Set orientation of time vs frequency axis. Can be @code{vertical} or
18072 @code{horizontal}. Default is @code{vertical}.
18075 Set scale gain for calculating intensity color values.
18076 Default value is @code{1}.
18079 Draw time and frequency axes and legends. Default is enabled.
18082 Set color rotation, must be in [-1.0, 1.0] range.
18083 Default value is @code{0}.
18086 @subsection Examples
18090 Extract an audio spectrogram of a whole audio track
18091 in a 1024x1024 picture using @command{ffmpeg}:
18093 ffmpeg -i audio.flac -lavfi showspectrumpic=s=1024x1024 spectrogram.png
18097 @section showvolume
18099 Convert input audio volume to a video output.
18101 The filter accepts the following options:
18108 Set border width, allowed range is [0, 5]. Default is 1.
18111 Set channel width, allowed range is [80, 8192]. Default is 400.
18114 Set channel height, allowed range is [1, 900]. Default is 20.
18117 Set fade, allowed range is [0.001, 1]. Default is 0.95.
18120 Set volume color expression.
18122 The expression can use the following variables:
18126 Current max volume of channel in dB.
18132 Current channel number, starting from 0.
18136 If set, displays channel names. Default is enabled.
18139 If set, displays volume values. Default is enabled.
18142 Set orientation, can be @code{horizontal} or @code{vertical},
18143 default is @code{horizontal}.
18146 Set step size, allowed range s [0, 5]. Default is 0, which means
18152 Convert input audio to a video output, representing the samples waves.
18154 The filter accepts the following options:
18158 Specify the video size for the output. For the syntax of this option, check the
18159 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18160 Default value is @code{600x240}.
18165 Available values are:
18168 Draw a point for each sample.
18171 Draw a vertical line for each sample.
18174 Draw a point for each sample and a line between them.
18177 Draw a centered vertical line for each sample.
18180 Default value is @code{point}.
18183 Set the number of samples which are printed on the same column. A
18184 larger value will decrease the frame rate. Must be a positive
18185 integer. This option can be set only if the value for @var{rate}
18186 is not explicitly specified.
18189 Set the (approximate) output frame rate. This is done by setting the
18190 option @var{n}. Default value is "25".
18192 @item split_channels
18193 Set if channels should be drawn separately or overlap. Default value is 0.
18196 Set colors separated by '|' which are going to be used for drawing of each channel.
18199 Set amplitude scale.
18201 Available values are:
18219 @subsection Examples
18223 Output the input file audio and the corresponding video representation
18226 amovie=a.mp3,asplit[out0],showwaves[out1]
18230 Create a synthetic signal and show it with showwaves, forcing a
18231 frame rate of 30 frames per second:
18233 aevalsrc=sin(1*2*PI*t)*sin(880*2*PI*t):cos(2*PI*200*t),asplit[out0],showwaves=r=30[out1]
18237 @section showwavespic
18239 Convert input audio to a single video frame, representing the samples waves.
18241 The filter accepts the following options:
18245 Specify the video size for the output. For the syntax of this option, check the
18246 @ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}.
18247 Default value is @code{600x240}.
18249 @item split_channels
18250 Set if channels should be drawn separately or overlap. Default value is 0.
18253 Set colors separated by '|' which are going to be used for drawing of each channel.
18256 Set amplitude scale.
18258 Available values are:
18276 @subsection Examples
18280 Extract a channel split representation of the wave form of a whole audio track
18281 in a 1024x800 picture using @command{ffmpeg}:
18283 ffmpeg -i audio.flac -lavfi showwavespic=split_channels=1:s=1024x800 waveform.png
18287 @section sidedata, asidedata
18289 Delete frame side data, or select frames based on it.
18291 This filter accepts the following options:
18295 Set mode of operation of the filter.
18297 Can be one of the following:
18301 Select every frame with side data of @code{type}.
18304 Delete side data of @code{type}. If @code{type} is not set, delete all side
18310 Set side data type used with all modes. Must be set for @code{select} mode. For
18311 the list of frame side data types, refer to the @code{AVFrameSideDataType} enum
18312 in @file{libavutil/frame.h}. For example, to choose
18313 @code{AV_FRAME_DATA_PANSCAN} side data, you must specify @code{PANSCAN}.
18317 @section spectrumsynth
18319 Sythesize audio from 2 input video spectrums, first input stream represents
18320 magnitude across time and second represents phase across time.
18321 The filter will transform from frequency domain as displayed in videos back
18322 to time domain as presented in audio output.
18324 This filter is primarily created for reversing processed @ref{showspectrum}
18325 filter outputs, but can synthesize sound from other spectrograms too.
18326 But in such case results are going to be poor if the phase data is not
18327 available, because in such cases phase data need to be recreated, usually
18328 its just recreated from random noise.
18329 For best results use gray only output (@code{channel} color mode in
18330 @ref{showspectrum} filter) and @code{log} scale for magnitude video and
18331 @code{lin} scale for phase video. To produce phase, for 2nd video, use
18332 @code{data} option. Inputs videos should generally use @code{fullframe}
18333 slide mode as that saves resources needed for decoding video.
18335 The filter accepts the following options:
18339 Specify sample rate of output audio, the sample rate of audio from which
18340 spectrum was generated may differ.
18343 Set number of channels represented in input video spectrums.
18346 Set scale which was used when generating magnitude input spectrum.
18347 Can be @code{lin} or @code{log}. Default is @code{log}.
18350 Set slide which was used when generating inputs spectrums.
18351 Can be @code{replace}, @code{scroll}, @code{fullframe} or @code{rscroll}.
18352 Default is @code{fullframe}.
18355 Set window function used for resynthesis.
18358 Set window overlap. In range @code{[0, 1]}. Default is @code{1},
18359 which means optimal overlap for selected window function will be picked.
18362 Set orientation of input videos. Can be @code{vertical} or @code{horizontal}.
18363 Default is @code{vertical}.
18366 @subsection Examples
18370 First create magnitude and phase videos from audio, assuming audio is stereo with 44100 sample rate,
18371 then resynthesize videos back to audio with spectrumsynth:
18373 ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=log:overlap=0.875:color=channel:slide=fullframe:data=magnitude -an -c:v rawvideo magnitude.nut
18374 ffmpeg -i input.flac -lavfi showspectrum=mode=separate:scale=lin:overlap=0.875:color=channel:slide=fullframe:data=phase -an -c:v rawvideo phase.nut
18375 ffmpeg -i magnitude.nut -i phase.nut -lavfi spectrumsynth=channels=2:sample_rate=44100:win_func=hann:overlap=0.875:slide=fullframe output.flac
18379 @section split, asplit
18381 Split input into several identical outputs.
18383 @code{asplit} works with audio input, @code{split} with video.
18385 The filter accepts a single parameter which specifies the number of outputs. If
18386 unspecified, it defaults to 2.
18388 @subsection Examples
18392 Create two separate outputs from the same input:
18394 [in] split [out0][out1]
18398 To create 3 or more outputs, you need to specify the number of
18401 [in] asplit=3 [out0][out1][out2]
18405 Create two separate outputs from the same input, one cropped and
18408 [in] split [splitout1][splitout2];
18409 [splitout1] crop=100:100:0:0 [cropout];
18410 [splitout2] pad=200:200:100:100 [padout];
18414 Create 5 copies of the input audio with @command{ffmpeg}:
18416 ffmpeg -i INPUT -filter_complex asplit=5 OUTPUT
18422 Receive commands sent through a libzmq client, and forward them to
18423 filters in the filtergraph.
18425 @code{zmq} and @code{azmq} work as a pass-through filters. @code{zmq}
18426 must be inserted between two video filters, @code{azmq} between two
18429 To enable these filters you need to install the libzmq library and
18430 headers and configure FFmpeg with @code{--enable-libzmq}.
18432 For more information about libzmq see:
18433 @url{http://www.zeromq.org/}
18435 The @code{zmq} and @code{azmq} filters work as a libzmq server, which
18436 receives messages sent through a network interface defined by the
18437 @option{bind_address} option.
18439 The received message must be in the form:
18441 @var{TARGET} @var{COMMAND} [@var{ARG}]
18444 @var{TARGET} specifies the target of the command, usually the name of
18445 the filter class or a specific filter instance name.
18447 @var{COMMAND} specifies the name of the command for the target filter.
18449 @var{ARG} is optional and specifies the optional argument list for the
18450 given @var{COMMAND}.
18452 Upon reception, the message is processed and the corresponding command
18453 is injected into the filtergraph. Depending on the result, the filter
18454 will send a reply to the client, adopting the format:
18456 @var{ERROR_CODE} @var{ERROR_REASON}
18460 @var{MESSAGE} is optional.
18462 @subsection Examples
18464 Look at @file{tools/zmqsend} for an example of a zmq client which can
18465 be used to send commands processed by these filters.
18467 Consider the following filtergraph generated by @command{ffplay}
18469 ffplay -dumpgraph 1 -f lavfi "
18470 color=s=100x100:c=red [l];
18471 color=s=100x100:c=blue [r];
18472 nullsrc=s=200x100, zmq [bg];
18473 [bg][l] overlay [bg+l];
18474 [bg+l][r] overlay=x=100 "
18477 To change the color of the left side of the video, the following
18478 command can be used:
18480 echo Parsed_color_0 c yellow | tools/zmqsend
18483 To change the right side:
18485 echo Parsed_color_1 c pink | tools/zmqsend
18488 @c man end MULTIMEDIA FILTERS
18490 @chapter Multimedia Sources
18491 @c man begin MULTIMEDIA SOURCES
18493 Below is a description of the currently available multimedia sources.
18497 This is the same as @ref{movie} source, except it selects an audio
18503 Read audio and/or video stream(s) from a movie container.
18505 It accepts the following parameters:
18509 The name of the resource to read (not necessarily a file; it can also be a
18510 device or a stream accessed through some protocol).
18512 @item format_name, f
18513 Specifies the format assumed for the movie to read, and can be either
18514 the name of a container or an input device. If not specified, the
18515 format is guessed from @var{movie_name} or by probing.
18517 @item seek_point, sp
18518 Specifies the seek point in seconds. The frames will be output
18519 starting from this seek point. The parameter is evaluated with
18520 @code{av_strtod}, so the numerical value may be suffixed by an IS
18521 postfix. The default value is "0".
18524 Specifies the streams to read. Several streams can be specified,
18525 separated by "+". The source will then have as many outputs, in the
18526 same order. The syntax is explained in the ``Stream specifiers''
18527 section in the ffmpeg manual. Two special names, "dv" and "da" specify
18528 respectively the default (best suited) video and audio stream. Default
18529 is "dv", or "da" if the filter is called as "amovie".
18531 @item stream_index, si
18532 Specifies the index of the video stream to read. If the value is -1,
18533 the most suitable video stream will be automatically selected. The default
18534 value is "-1". Deprecated. If the filter is called "amovie", it will select
18535 audio instead of video.
18538 Specifies how many times to read the stream in sequence.
18539 If the value is 0, the stream will be looped infinitely.
18540 Default value is "1".
18542 Note that when the movie is looped the source timestamps are not
18543 changed, so it will generate non monotonically increasing timestamps.
18545 @item discontinuity
18546 Specifies the time difference between frames above which the point is
18547 considered a timestamp discontinuity which is removed by adjusting the later
18551 It allows overlaying a second video on top of the main input of
18552 a filtergraph, as shown in this graph:
18554 input -----------> deltapts0 --> overlay --> output
18557 movie --> scale--> deltapts1 -------+
18559 @subsection Examples
18563 Skip 3.2 seconds from the start of the AVI file in.avi, and overlay it
18564 on top of the input labelled "in":
18566 movie=in.avi:seek_point=3.2, scale=180:-1, setpts=PTS-STARTPTS [over];
18567 [in] setpts=PTS-STARTPTS [main];
18568 [main][over] overlay=16:16 [out]
18572 Read from a video4linux2 device, and overlay it on top of the input
18575 movie=/dev/video0:f=video4linux2, scale=180:-1, setpts=PTS-STARTPTS [over];
18576 [in] setpts=PTS-STARTPTS [main];
18577 [main][over] overlay=16:16 [out]
18581 Read the first video stream and the audio stream with id 0x81 from
18582 dvd.vob; the video is connected to the pad named "video" and the audio is
18583 connected to the pad named "audio":
18585 movie=dvd.vob:s=v:0+#0x81 [video] [audio]
18589 @subsection Commands
18591 Both movie and amovie support the following commands:
18594 Perform seek using "av_seek_frame".
18595 The syntax is: seek @var{stream_index}|@var{timestamp}|@var{flags}
18598 @var{stream_index}: If stream_index is -1, a default
18599 stream is selected, and @var{timestamp} is automatically converted
18600 from AV_TIME_BASE units to the stream specific time_base.
18602 @var{timestamp}: Timestamp in AVStream.time_base units
18603 or, if no stream is specified, in AV_TIME_BASE units.
18605 @var{flags}: Flags which select direction and seeking mode.
18609 Get movie duration in AV_TIME_BASE units.
18613 @c man end MULTIMEDIA SOURCES